Systems and methods for mentor-mentee matching

Abstract

Systems and methods for enabling pairing between one or more first entities and one or more second entities are disclosed. In an aspect, the one or more first entities may be mentees and the one or more second entities may be mentors. Each mentee and each mentor may be represented by a plurality of vectors having elements representing characteristics relevant to pairing. One or more metrics of similarity and / or difference for a potential mentee and mentor pair may be determined based on one or more vector operations, and these metrics may be ranked from the perspective of the mentee and the perspective of the mentor. A global ranking metric value may be determined for each mentee-mentor pair, and a mentee may be matched with a mentor based on the global ranking metric value.

Description

SYSTEMS AND METHODS FOR MENTOR-MENTEE MATCHINGCROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the priority benefit, under 35 U.S.C. 119(e), of U.S. Application No. 63 / 610,312, filed December 14, 2023 and entitled “Systems and Methods for Mentor-Mentee Matching,” and U.S. Application No. 63 / 682,634, filed August 13, 2024 and entitled “Systems and Methods for Mentor-Mentee Matching,” both of which are incorporated by reference in their entirety for all purposes.BACKGROUND

[0002] Employees early in their careers may benefit from mentorship from more senior or seasoned employees who have gained valuable experience over the course of their careers. For large organizations, matching mentors and mentees can be difficult given the number of potential mentors and mentees as well as a variety of constraints and / or criteria for the matching process. Whether or not a mentor is a good match for a mentee may depend on a wide variety of factors. Few, if any, solutions for efficiently and effectively matching mentors with mentees exist for large entities such as corporations. Further, human-mediated matching processes may raise questions of favoritism and bias and undermine the effectiveness of the generated matches.SUMMARY

[0003] Mentorship can be a critical part of an employee’s career and professional skillset development. However, conventional solutions for matching mentors and mentees can be slow, unsophisticated, prone to human bias, and can fail to account for a variety of parameters that may have significant bearing on the outcome of the mentorship process, particularly when those parameters change over time or are particular to an individual employee.

[0004] Accordingly, the present disclosure is directed toward providing improved solutions and inventive methods for mentor-mentee matching. The matches provided by the present disclosure may be based on information including preferences and parameters provided byboth mentors and mentees, and may enable a rapid, scalable, objective and unbiased pairing of mentors with mentees based on a variety of relevant considerations and criteria for matching. Further, the present disclosure may significantly reduce the risk of favoritism introduced by human-managed methods.

[0005] The techniques described herein are especially valuable in light of Moravec’s Paradox, in which tasks that are easy for computers to perform are difficult for humans to perform, and vice versa. In this case, the present techniques provide systems and methods for rapidly performing large-scale, unbiased, complex, and interconnected matching for hundreds or thousands of pairs of entities. In particular, no human-implemented method attempting to mimic the present disclosure could completely eliminate inherent human bias in a process for matching other humans, nor would such a human-implemented method be able to perform the requisite large-scale calculations required to process entity matches across an organization such as a corporation with tens of thousands of employees or a candidate therapeutic drug with potentially millions of compatible biomarkers.

[0006] More generally, the inventive concepts disclosed herein are more broadly applicable to matching two or more entities where each entity provides (or is described by) entityspecific parameters or guidelines (e.g., a wish list) and only one entity provides (or is described by) an additional filter indicating a preference, characteristic or affinity toward one or more of the provided parameters. For example, the inventive concepts disclosed herein are applicable to recruitment for jobs and / or volunteering (e.g., where job applicants are analogous to mentors and open positions are analogous to mentees), education (e.g., where students are analogous to mentees and teachers / tutors / academic advisors / research advisors are analogous to mentors), healthcare (e.g., where patients are analogous to mentees and providers / specialists are analogous to mentors), customer support (e.g., where customers are analogous to mentees and customer service representatives are analogous to mentors), or any suitable field.

[0007] In some aspects, the techniques described herein relate to a method for creating mentor-mentee matches between a plurality of mentors and a plurality of mentees, each mentor of the plurality of mentors associated with a corresponding n-element mentor skill goal vector and a corresponding m-element mentor activity goal vector, and each mentee of the plurality of mentees associated with a corresponding n-element mentee skill goal vector and a corresponding m-element mentee activity goal vector, the method including: (a) generating a mentee perspective pair array, the mentee perspective pair array including theplurality of mentees and a plurality of eligible and compatible mentors of the plurality of mentors represented as pairs; (b) identifying a selected mentee of the plurality of mentees; (c) calculating, for each pair of the mentee perspective pair array including the selected mentee, a dot product of an n-element mentee skill goal vector for the selected mentee and an n-element mentor skill goal vector for the mentor of that pair to generate a skill similarity for that pair; (d) calculating, for each pair of the mentee perspective pair array including the selected mentee, a Euclidean distance of an m-element mentee activity goal vector for the selected mentee and an m-element mentor activity goal vector for the mentor of that pair to generate an activity distance for that pair; (e) generating, for each pair of the mentee perspective pair array including the selected mentee, a ranking pair, the ranking pair including a ranking of the skill similarity and the activity distance for that pair of the mentee perspective pair array, and placing the ranking pair into a mentee ranking group; (f) calculating, for each pair of the mentee ranking group and based on the ranking of the skill similarity and the activity distance for that pair, a merged mentee perspective rank for that pair; (g) repeating B through F for each additional mentee of the plurality of mentees to generate a merged mentee perspective rank for each possible pair of mentees and mentors; (h) generating a mentor perspective pair array, the mentor perspective pair array including the plurality of eligible and compatible mentors of the plurality of mentors and the plurality of mentees represented as pairs; (i) identifying a selected mentor of the plurality of eligible and compatible mentors; (j) calculating, for each pair of the mentor perspective pair array including the selected mentor, a dot product of an n-element mentor skill goal vector for the selected mentor and an n-element mentee skill goal vector for the mentee of that pair to generate a skill similarity for that pair; (k) calculating, for each pair of the mentor perspective pair array including the selected mentor, a Euclidean distance of an m-element mentor activity goal vector for the selected mentor and an m-element mentee activity goal vector for the mentee of that pair to generate an activity distance for that pair; (1) generating, for each pair of the mentor perspective pair array including the selected mentor, a ranking pair, the ranking pair including a ranking of the skill similarity and the activity distance for that pair of the mentor perspective pair array, and placing the ranking pair into a mentor ranking group; (m) calculating, for each pair of the mentor ranking group and based on the ranking of the skill similarity and the activity distance for that pair, a merged mentor perspective rank for that pair; (n) repeating I through M for each additional mentor of the plurality of eligible and compatible mentors to generate a merged mentor perspective rank for each possible pair of mentors and mentees; (o) combining, for each possible mentor-mentee pair from the plurality of mentees and theplurality of eligible and compatible mentors, the merged mentor perspective rank and the merged mentee perspective rank to determine a global ranking metric for that mentor-mentee pair; and (p) matching a mentee of the plurality of mentees with a mentor of the plurality of eligible and compatible mentors based on the global ranking metric.

[0008] In some aspects, the techniques described herein relate to a method, wherein P) further includes at least one of: transmitting a notification to an approver that a potential pair match between the mentee of the plurality of mentees and the mentor of the eligible and compatible mentors requires approval; and transmitting a notification to the mentee of the plurality of mentees indicating that they have been matched with the mentor of the plurality of eligible and compatible mentors.

[0009] In some aspects, the techniques described herein relate to a method, further including:Q) in response to the mentee not responding to the notification within a predetermined time limit, deleting the match between the mentee and the mentor and making the mentee and the mentor available for matching again.

[0010] In some aspects, the techniques described herein relate to a method, further including: performing a plurality of B) - F) and I) - M) in parallel, wherein performing a plurality of B) - F) and I) - M) in parallel includes reducing a time to complete the plurality of B) - F) and I) - M) by distributing at least a portion of B) - F) and I) - M) across a plurality of processors or a plurality of cores; and P) further includes asynchronously presenting a plurality of administrators with data associated with matching the mentee of the plurality of mentees and the mentor of the plurality of eligible and compatible mentors and receiving an approval, by at least one of the administrators, of the match between the mentee of the plurality of mentees and the mentor of the plurality of eligible and compatible mentors.

[0011] In some aspects, the techniques described herein relate to a method, further including:R) generating, using a large language model, at least one recommendation for a discussion topic between the mentee of the plurality of mentees and the mentor of the eligible and compatible mentors based on information provided by the mentee and the mentor; and S) transmitting, to at least one of the mentee of the plurality of mentees or the mentor of the eligible and compatible mentors, the at least one recommendation for a discussion topic.

[0012] In some aspects, the techniques described herein relate to a method, further including electronically scheduling a meeting between the matched mentee of the plurality of mentees and the matched mentor of the plurality of eligible and compatible mentors.

[0013] In some aspects, the techniques described herein relate to a method, wherein electronically scheduling a meeting includes transmitting an indication of the meeting to the matched mentee of the plurality of mentees and the matched mentor of the plurality of eligible and compatible mentors.

[0014] In some aspects, the techniques described herein relate to a method, wherein electronically scheduling a meeting further includes restricting an access of a physical location based on the matching.

[0015] In some aspects, the techniques described herein relate to a method, wherein restricting the access of the physical location includes controlling a lock.

[0016] In some aspects, the techniques described herein relate to a method, wherein: the n- element mentor skill goal vectors include n mentor skill elements, the n mentor skill elements including values between 0 and 1; and the n-element mentee skill goal vectors include n mentee skill elements, the n mentee skill elements including integer values from 1 to n.

[0017] In some aspects, the techniques described herein relate to a method, wherein: the m- element mentor activity goal vectors include m mentor activity elements, the m mentor activity elements including integer values from 1 to m; and the m-element mentee activity goal vectors include m mentee activity elements, the m mentee activity elements including integer values from 1 to m.

[0018] In some aspects, the techniques described herein relate to a method, further including: Q) receiving, by a generative artificial intelligence (Al) model, first information from the mentee of the plurality of mentees or the mentor of the plurality of mentors; and R) outputting, by the generative Al model, second information in response to the first information.

[0019] In some aspects, the techniques described herein relate to a method, wherein the first information includes a prompt.

[0020] In some aspects, the techniques described herein relate to a method, wherein the prompt includes a natural language request for information associated with human conversation.

[0021] In some aspects, the techniques described herein relate to a method, wherein the information associated with human conversation includes one or more suggestions for discussion topics.

[0022] In some aspects, the techniques described herein relate to a method, wherein the second information includes one or more natural language suggestions for discussion topics.

[0023] In some aspects, the techniques described herein relate to a method, wherein the generative Al model includes at least one of a large language model (LLM), a generative pretrained transformer (GPT), a generative adversarial network (GAN), a neural network, or a diffusion model.

[0024] In some aspects, the techniques described herein relate to a method, wherein: the generative Al model utilizes retrieval augmented generation (RAG) to supplement the first information with third information, the third information received from a source other than the mentee of the plurality of mentees or the mentor of the plurality of mentors; and the generative Al model generates the second information based on the first information and the third information.

[0025] In some aspects, the techniques described herein relate to a method, wherein the second information is based on at least one of personal information associated with the mentee or personal information associated with the mentor.

[0026] In some aspects, the techniques described herein relate to a method, wherein the second information is based on at least one category of shared interests between the mentee and the mentor.

[0027] In some aspects, the techniques described herein relate to a method, wherein the second information includes a natural language output.

[0028] In some aspects, the techniques described herein relate to a non-transitory computer- readable medium containing instructions, the instructions configuring a processor to execute a method for creating mentor-mentee matches between a plurality of mentors and a plurality of mentees, each mentor of the plurality of mentors associated with a corresponding n- element mentor skill goal vector and a corresponding m-element mentor activity goal vector, and each mentee of the plurality of mentees associated with a corresponding n-element mentee skill goal vector and a corresponding m-element mentee activity goal vector, the method including: (a) generating a mentee perspective pair array, the mentee perspective pair array including the plurality of mentees and a plurality of eligible and compatible mentors of the plurality of mentors represented as pairs; (b) identifying a selected mentee of the plurality of mentees; (c) calculating, for each pair of the mentee perspective pair array including the selected mentee, a dot product of an n-element mentee skill goal vector for the selectedmentee and an n-element mentor skill goal vector for the mentor of that pair to generate a skill similarity for that pair; (d) calculating, for each pair of the mentee perspective pair array including the selected mentee, a Euclidean distance of an m-element mentee activity goal vector for the selected mentee and an m-element mentor activity goal vector for the mentor of that pair to generate an activity distance for that pair; (e) generating, for each pair of the mentee perspective pair array including the selected mentee, a ranking pair, the ranking pair including a ranking of the skill similarity and the activity distance for that pair of the mentee perspective pair array, and placing the ranking pair into a mentee ranking group; (f) calculating, for each pair of the mentee ranking group and based on the ranking of the skill similarity and the activity distance for that pair, a merged mentee perspective rank for that pair; (g) repeating B through F for each additional mentee of the plurality of mentees to generate a merged mentee perspective rank for each possible pair of mentees and mentors; (h) generating a mentor perspective pair array, the mentor perspective pair array including the plurality of eligible mentors of the plurality of mentors and the plurality of mentees represented as pairs; (i) identifying, based on one or more second eligibility criteria, a selected mentor of the plurality of eligible mentors; (j) calculating, for each pair of the mentor perspective pair array including the selected mentor, a dot product of an n-element mentor skill goal vector for the selected mentor and an n-element mentee skill goal vector for the mentee of that pair to generate a skill similarity for that pair; (k) calculating, for each pair of the mentor perspective pair array including the selected mentor, a Euclidean distance of an m-element mentor activity goal vector for the selected mentor and an m-element mentee activity goal vector for the mentee of that pair to generate an activity distance for that pair; (1) generating, for each pair of the mentor perspective pair array including the selected mentor, a ranking pair, the ranking pair including a ranking of the skill similarity and the activity distance for that pair of the mentor perspective pair array, and placing the ranking pair into a mentor ranking group; (m) calculating, for each pair of the mentor ranking group and based on the ranking of the skill similarity and the activity distance for that pair, a merged mentor perspective rank for that pair; (n) repeating I through M for each additional mentor of the plurality of eligible mentors to generate a merged mentor perspective rank for each possible pair of mentors and mentees; (o) combining, for each possible mentor-mentee pair from the plurality of mentees and the plurality of eligible mentors, the merged mentor perspective rank and the merged mentee perspective rank to determine a global ranking metric for that mentor- mentee pair; and (p) matching a mentee of the plurality of mentees with a mentor of the plurality of eligible mentors based on the global ranking metric.

[0029] In some aspects, the techniques described herein relate to a medium, wherein P) further includes at least one of: transmitting a notification to an approver that a potential pair match between the mentee of the plurality of mentees and the mentor of the eligible and compatible mentors requires approval; and transmitting a notification to the mentee of the plurality of mentees indicating that they have been matched with the mentor of the plurality of eligible and compatible mentors.

[0030] In some aspects, the techniques described herein relate to a medium, further including:Q) in response to the mentee not responding to the notification within a predetermined time limit, deleting the match between the mentee and the mentor and making the mentee and the mentor available for matching again.

[0031] In some aspects, the techniques described herein relate to a medium, further including: performing a plurality of B) - F) and I) - M) in parallel, wherein performing a plurality of B) - F) and I) - M) in parallel includes reducing a time to complete the plurality of B) - F) and I) - M) by distributing at least a portion of B) - F) and I) - M) across a plurality of processors or a plurality of cores; and P) further includes asynchronously presenting a plurality of administrators with data associated with matching the mentee of the plurality of mentees and the mentor of the plurality of eligible and compatible mentors and receiving an approval, by at least one of the administrators, of the match between the mentee of the plurality of mentees and the mentor of the plurality of eligible and compatible mentors.

[0032] In some aspects, the techniques described herein relate to a medium, further including:R) generating, using a large language model, at least one recommendation for a discussion topic between the mentee of the plurality of mentees and the mentor of the eligible and compatible mentors based on information provided by the mentee and the mentor; and S) transmitting, to at least one of the mentee of the plurality of mentees or the mentor of the eligible and compatible mentors, the at least one recommendation for a discussion topic.

[0033] In some aspects, the techniques described herein relate to a medium, further including electronically scheduling a meeting between the matched mentee of the plurality of mentees and the matched mentor of the plurality of eligible and compatible mentors.

[0034] In some aspects, the techniques described herein relate to a medium, wherein electronically scheduling a meeting includes transmitting an indication of the meeting to the matched mentee of the plurality of mentees and the matched mentor of the plurality of eligible and compatible mentors.

[0035] In some aspects, the techniques described herein relate to a medium, wherein electronically scheduling a meeting further includes restricting an access of a physical location based on the matching.

[0036] In some aspects, the techniques described herein relate to a medium, wherein restricting the access of the physical location includes controlling a lock.

[0037] In some aspects, the techniques described herein relate to a medium, wherein: the n- element mentor skill goal vectors include n mentor skill elements, the n mentor skill elements including values between 0 and 1; and the n-element mentee skill goal vectors include n mentee skill elements, the n mentee skill elements including integer values from 1 to n.

[0038] In some aspects, the techniques described herein relate to a medium, wherein: the m- element mentor activity goal vectors include m mentor activity elements, the m mentor activity elements including integer values from 1 to m; and the m-element mentee activity goal vectors include m mentee activity elements, the m mentee activity elements including integer values from 1 to m.

[0039] In some aspects, the techniques described herein relate to a medium, further including: Q) receiving, by a generative artificial intelligence (Al) model, first information from the mentee of the plurality of mentees or the mentor of the plurality of mentors; and R) outputting, by the generative Al model, second information in response to the first information.

[0040] In some aspects, the techniques described herein relate to a medium, wherein the first information includes a prompt.

[0041] In some aspects, the techniques described herein relate to a medium, wherein the prompt includes a natural language request for information associated with human conversation.

[0042] In some aspects, the techniques described herein relate to a medium, wherein the information associated with human conversation includes one or more suggestions for discussion topics.

[0043] In some aspects, the techniques described herein relate to a medium, wherein the second information includes one or more natural language suggestions for discussion topics.

[0044] In some aspects, the techniques described herein relate to a medium, wherein the generative Al model includes at least one of a large language model (LLM), a generative pre-trained transformer (GPT), a generative adversarial network (GAN), a neural network, or a diffusion model.

[0045] In some aspects, the techniques described herein relate to a medium, wherein: the generative Al model utilizes retrieval augmented generation (RAG) to supplement the first information with third information, the third information received from a source other than the mentee of the plurality of mentees or the mentor of the plurality of mentors; and the generative Al model generates the second information based on the first information and the third information.

[0046] In some aspects, the techniques described herein relate to a medium, wherein the second information is based on at least one of personal information associated with the mentee or personal information associated with the mentor.

[0047] In some aspects, the techniques described herein relate to a medium, wherein the second information is based on at least one category of shared interests between the mentee and the mentor.

[0048] In some aspects, the techniques described herein relate to a medium, wherein the second information includes a natural language output.

[0049] In some aspects, the techniques described herein relate to a system for creating mentor-mentee matches between a plurality of mentors and a plurality of mentees, each mentor of the plurality of mentors associated with a corresponding n-element mentor skill goal vector and a corresponding m-element mentor activity goal vector, and each mentee of the plurality of mentees associated with a corresponding n-element mentee skill goal vector and a corresponding m-element mentee activity goal vector, the system including: a processor; and a memory, the memory including instructions configuring the processor to execute a method, the method including: (a) generating a mentee perspective pair array, the mentee perspective pair array including the plurality of mentees and a plurality of eligible and compatible mentors of the plurality of mentors represented as pairs; (b) identifying a selected mentee of the plurality of mentees; (c) calculating, for each pair of the mentee perspective pair array including the selected mentee, a dot product of an n-element mentee skill goal vector for the selected mentee and an n-element mentor skill goal vector for the mentor of that pair to generate a skill similarity for that pair; (d) calculating, for each pair of the mentee perspective pair array including the selected mentee, a Euclidean distance of an m-element mentee activity goal vector for the selected mentee and an m-element mentor activity goalvector for the mentor of that pair to generate an activity distance for that pair; (e) generating, for each pair of the mentee perspective pair array including the selected mentee, a ranking pair, the ranking pair including a ranking of the skill similarity and the activity distance for that pair of the mentee perspective pair array, and placing the ranking pair into a mentee ranking group; (f) calculating, for each pair of the mentee ranking group and based on the ranking of the skill similarity and the activity distance for that pair, a merged mentee perspective rank for that pair; (g) repeating B through F for each additional mentee of the plurality of mentees to generate a merged mentee perspective rank for each possible pair of mentees and mentors; (h) generating a mentor perspective pair array, the mentor perspective pair array including the plurality of eligible and compatible mentors of the plurality of mentors and the plurality of mentees represented as pairs; (i) identifying a selected mentor of the plurality of eligible and compatible mentors; (j) calculating, for each pair of the mentor perspective pair array including the selected mentor, a dot product of an n-element mentor skill goal vector for the selected mentor and an n-element mentee skill goal vector for the mentee of that pair to generate a skill similarity for that pair; (k) calculating, for each pair of the mentor perspective pair array including the selected mentor, a Euclidean distance of an m-element mentor activity goal vector for the selected mentor and an m-element mentee activity goal vector for the mentee of that pair to generate an activity distance for that pair; (1) generating, for each pair of the mentor perspective pair array including the selected mentor, a ranking pair, the ranking pair including a ranking of the skill similarity and the activity distance for that pair of the mentor perspective pair array, and placing the ranking pair into a mentor ranking group; (m) calculating, for each pair of the mentor ranking group and based on the ranking of the skill similarity and the activity distance for that pair, a merged mentor perspective rank for that pair; (n) repeating I through M for each additional mentor of the plurality of eligible and compatible mentors to generate a merged mentor perspective rank for each possible pair of mentors and mentees; (o) combining, for each possible mentor-mentee pair from the plurality of mentees and the plurality of eligible and compatible mentors, the merged mentor perspective rank and the merged mentee perspective rank to determine a global ranking metric for that mentor-mentee pair; and (p) matching a mentee of the plurality of mentees with a mentor of the plurality of eligible and compatible mentors based on the global ranking metric.

[0050] In some aspects, the techniques described herein relate to a system, wherein P) further includes at least one of: transmitting a notification to an approver that a potential pair matchbetween the mentee of the plurality of mentees and the mentor of the eligible and compatible mentors requires approval; and transmitting a notification to the mentee of the plurality of mentees indicating that they have been matched with the mentor of the plurality of eligible and compatible mentors.

[0051] In some aspects, the techniques described herein relate to a system, further including:Q) in response to the mentee not responding to the notification within a predetermined time limit, deleting the match between the mentee and the mentor and making the mentee and the mentor available for matching again.

[0052] In some aspects, the techniques described herein relate to a system, further including: performing a plurality of B) - F) and I) - M) in parallel, wherein performing a plurality of B) - F) and I) - M) in parallel includes reducing a time to complete the plurality of B) - F) and I) - M) by distributing at least a portion of B) - F) and I) - M) across a plurality of processors or a plurality of cores; and P) further includes asynchronously presenting a plurality of administrators with data associated with matching the mentee of the plurality of mentees and the mentor of the plurality of eligible and compatible mentors and receiving an approval, by at least one of the administrators, of the match between the mentee of the plurality of mentees and the mentor of the plurality of eligible and compatible mentors.

[0053] In some aspects, the techniques described herein relate to a system, further including:R) generating, using a large language model, at least one recommendation for a discussion topic between the mentee of the plurality of mentees and the mentor of the eligible and compatible mentors based on information provided by the mentee and the mentor; and S) transmitting, to at least one of the mentee of the plurality of mentees or the mentor of the eligible and compatible mentors, the at least one recommendation for a discussion topic.

[0054] In some aspects, the techniques described herein relate to a system, further including electronically scheduling a meeting between the matched mentee of the plurality of mentees and the matched mentor of the plurality of eligible and compatible mentors.

[0055] In some aspects, the techniques described herein relate to a system, wherein electronically scheduling a meeting includes transmitting an indication of the meeting to the matched mentee of the plurality of mentees and the matched mentor of the plurality of eligible and compatible mentors.

[0056] In some aspects, the techniques described herein relate to a system, wherein electronically scheduling a meeting further includes restricting an access of a physical location based on the matching.

[0057] In some aspects, the techniques described herein relate to a system, wherein restricting the access of the physical location includes controlling a lock.

[0058] In some aspects, the techniques described herein relate to a system, wherein: the n- element mentor skill goal vectors include n mentor skill elements, the n mentor skill elements including values between 0 and 1; and the n-element mentee skill goal vectors include n mentee skill elements, the n mentee skill elements including integer values from 1 to n.

[0059] In some aspects, the techniques described herein relate to a system, wherein: the m- element mentor activity goal vectors include m mentor activity elements, the m mentor activity elements including integer values from 1 to m; and the m-element mentee activity goal vectors include m mentee activity elements, the m mentee activity elements including integer values from 1 to m.

[0060] In some aspects, the techniques described herein relate to a system, further including: Q) receiving, by a generative artificial intelligence (Al) model, first information from the mentee of the plurality of mentees or the mentor of the plurality of mentors; and R) outputting, by the generative Al model, second information in response to the first information.

[0061] In some aspects, the techniques described herein relate to a system, wherein the first information includes a prompt.

[0062] In some aspects, the techniques described herein relate to a system, wherein the prompt includes a natural language request for information associated with human conversation.

[0063] In some aspects, the techniques described herein relate to a system, wherein the information associated with human conversation includes one or more suggestions for discussion topics.

[0064] In some aspects, the techniques described herein relate to a system, wherein the second information includes one or more natural language suggestions for discussion topics.

[0065] In some aspects, the techniques described herein relate to a system, wherein the generative Al model includes at least one of a large language model (LLM), a generative pre-trained transformer (GPT), a generative adversarial network (GAN), a neural network, or a diffusion model.

[0066] In some aspects, the techniques described herein relate to a system, wherein: the generative Al model utilizes retrieval augmented generation (RAG) to supplement the first information with third information, the third information received from a source other than the mentee of the plurality of mentees or the mentor of the plurality of mentors; and the generative Al model generates the second information based on the first information and the third information.

[0067] In some aspects, the techniques described herein relate to a system, wherein the second information is based on at least one of personal information associated with the mentee or personal information associated with the mentor.

[0068] In some aspects, the techniques described herein relate to a system, wherein the second information is based on at least one category of shared interests between the mentee and the mentor.

[0069] In some aspects, the techniques described herein relate to a system, wherein the second information includes a natural language output.

[0070] In some aspects, the techniques described herein relate to a method for matching a plurality of first entities and a plurality of second entities, the method including: (a) calculating, for each potential pair including a selected first entity of the plurality of first entities and a selected second entity of the plurality of second entities, a similarity metric and a distance metric; (b) ranking each potential pair relative to all other potential pairs based on the similarity metric and the distance metric for that potential pair to generate a list of ranked potential pairs; (c) nominating a top ranked pair of the list of ranked potential pairs to be matched; (d) locking any of the other potential pairs from the list of ranked potential pairs including an entity whose quota for matches has been reached to generate a list of remaining ranked potential pairs; and (e) repeating C) and D) using the list of remaining ranked potential pairs until each of the plurality of first entities or each of the plurality of second entities has been matched.

[0071] In some aspects, the techniques described herein relate to a method, further including prior to A): receiving, from at least one entity of the plurality of first entities, first information associated with the at least one entity; inputting the first information associated with the at least one entity into a generative artificial intelligence (Al) model; generating, by thegenerative Al model, one or more responses indicating second information associated with at least one of an improved similarity metric or an improved distance metric for at least one potential pair including the at least one entity; and providing the second information to the at least one entity.

[0072] In some aspects, the techniques described herein relate to a method, wherein the generative Al model includes at least one of a large language model (LLM), a generative pretrained transformer (GPT), a generative adversarial network (GAN), a neural network, or a diffusion model.

[0073] In some aspects, the techniques described herein relate to a method, wherein inputting the first information into a generative Al model includes providing a prompt to the generative Al model.

[0074] In some aspects, the techniques described herein relate to a method, wherein the second information includes one or more suggestions for the at least one entity.

[0075] In some aspects, the techniques described herein relate to a method, wherein the one or more suggestions include a recommendation for organizing entity-specific information used during A).

[0076] In some aspects, the techniques described herein relate to a method, wherein generating, by the generative Al model, the one or more responses includes comparing the second information to a predetermined format.

[0077] In some aspects, the techniques described herein relate to a method, further including upon determining the one or more responses do not match the predetermined format, resubmitting the first information to the generative Al model.

[0078] In some aspects, the techniques described herein relate to a method, wherein: the generative Al model utilizes retrieval augmented generation (RAG) to supplement the first information with third information, the third information received from a source other than the first entity; and the generative Al model generates the second information based on the first information and the third information.

[0079] In some aspects, the techniques described herein relate to a method, wherein E) further includes at least one of: transmitting a notification to an approver that a potential pair requires approval; and transmitting a notification to the first entity of each potential pair indicating that they have been matched with the second entity.

[0080] In some aspects, the techniques described herein relate to a method, further including:F) in response to the first entity not responding to the notification within a predetermined time limit, forfeiting the match between the first entity and the second entity.

[0081] In some aspects, the techniques described herein relate to a method, further including: performing A) in parallel, wherein performing A) in parallel includes reducing a time to complete A) by distributing at least a portion of A) across a plurality of processors or a plurality of cores; and E) further includes asynchronously presenting a plurality of administrators with data associated with matching the first entity and the second entity and receiving an approval, by at least one of the administrators, of the match between the first entity and the second entity.

[0082] In some aspects, the techniques described herein relate to a method, further including:G) generating, using a large language model, at least one recommendation for a discussion topic between the first entity and the second entity based on information provided by the first entity and the second entity; and H) transmitting, to at least one of the first entity or the second entity, the at least one recommendation for a discussion topic.

[0083] In some aspects, the techniques described herein relate to a method, wherein the similarity metric is calculated based on an average, a dot product, a variance, a standard deviation, subtraction, or addition.

[0084] In some aspects, the techniques described herein relate to a method, wherein the distance metric is calculated based on a Euclidean distance, subtraction, addition, multiplication, a variance, or a standard deviation.

[0085] In some aspects, the techniques described herein relate to a method, wherein the selected first entity of the plurality of first entities includes: an n-element mentee skill goal vector including n mentee skill elements, the n mentee skill elements including integer values from 1 to n; and an m-element mentee activity goal vector including m mentee activity elements, the m mentee activity elements including integer values from 1 to m; and wherein the selected second entity of the plurality of second entities includes: an n-element mentor skill goal vector including n mentor skill elements, the n mentor skill elements including values between 0 and 1; and an m-element mentor activity goal vector including m mentor activity elements, the m mentor activity elements including integer values from 1 to m.

[0086] In some aspects, the techniques described herein relate to a method, further including electronically scheduling a meeting between each pair of matched entities.

[0087] In some aspects, the techniques described herein relate to a method, wherein electronically scheduling a meeting includes transmitting an indication of the meeting to each entity of each pair of matched entities.

[0088] In some aspects, the techniques described herein relate to a method, wherein electronically scheduling a meeting further includes restricting an access of a physical location based on the matching.

[0089] In some aspects, the techniques described herein relate to a method, wherein restricting the access of the physical location includes controlling a lock.

[0090] In some aspects, the techniques described herein relate to a non-transitory computer- readable medium containing instructions, the instructions configuring a processor to execute a method for matching a plurality of first entities and a plurality of second entities, the method including: (a) calculating, for each potential pair including a selected first entity of the plurality of first entities and a selected second entity of the plurality of second entities, a similarity metric and a distance metric; (b) ranking each potential pair relative to all other potential pairs based on the similarity metric and the distance metric for that potential pair to generate a list of ranked potential pairs; (c) nominating a top ranked pair of the list of ranked potential pairs to be matched; (d) locking any of the other potential pairs from the list of ranked potential pairs including an entity whose quota for matches has been reached to generate a list of remaining ranked potential pairs; and (e) repeating C) and D) using the list of remaining ranked potential pairs until each of the plurality of first entities or each of the plurality of second entities has been matched.

[0091] In some aspects, the techniques described herein relate to a medium, further including prior to A): receiving, from at least one entity of the plurality of first entities, first information associated with the at least one entity; inputting the first information associated with the at least one entity into a generative artificial intelligence (Al) model; generating, by the generative Al model, one or more responses indicating second information associated with at least one of an improved similarity metric or an improved distance metric for at least one potential pair including the at least one entity; and providing the second information to the at least one entity.

[0092] In some aspects, the techniques described herein relate to a medium, wherein the generative Al model includes at least one of a large language model (LLM), a generative pre-trained transformer (GPT), a generative adversarial network (GAN), a neural network, or a diffusion model.

[0093] In some aspects, the techniques described herein relate to a medium, wherein inputting the first information into a generative Al model includes providing a prompt to the generative Al model.

[0094] In some aspects, the techniques described herein relate to a medium, wherein the second information includes one or more suggestions for the at least one entity.

[0095] In some aspects, the techniques described herein relate to a medium, wherein the one or more suggestions include a recommendation for organizing entity-specific information used during A).

[0096] In some aspects, the techniques described herein relate to a medium, wherein generating, by the generative Al model, the one or more responses includes comparing the second information to a predetermined format.

[0097] In some aspects, the techniques described herein relate to a medium, further including upon determining the one or more responses do not match the predetermined format, resubmitting the first information to the generative Al model.

[0098] In some aspects, the techniques described herein relate to a medium, wherein: the generative Al model utilizes retrieval augmented generation (RAG) to supplement the first information with third information, the third information received from a source other than the first entity; and the generative Al model generates the second information based on the first information and the third information.

[0099] In some aspects, the techniques described herein relate to a medium, wherein E) further includes at least one of: transmitting a notification to an approver that a potential pair requires approval; and transmitting a notification to the first entity of each potential pair indicating that they have been matched with the second entity.

[0100] In some aspects, the techniques described herein relate to a medium, further including: F) in response to the first entity not responding to the notification within a predetermined time limit, deleting the match between the first entity and the second entity and making the first entity and the second entity available for matching again.

[0101] In some aspects, the techniques described herein relate to a medium, further including: performing A) in parallel, wherein performing A) in parallel includes reducing atime to complete A) by distributing at least a portion of A) across a plurality of processors or a plurality of cores; and E) further includes asynchronously presenting a plurality of administrators with data associated with matching the first entity and the second entity and receiving an approval, by at least one of the administrators, of the match between the first entity and the second entity.

[0102] In some aspects, the techniques described herein relate to a medium, further including: G) generating, using a large language model, at least one recommendation for a discussion topic between the first entity and the second entity based on information provided by the first entity and the second entity; and H) transmitting, to at least one of the first entity or the second entity, the at least one recommendation for a discussion topic.

[0103] In some aspects, the techniques described herein relate to a medium, wherein the similarity metric is calculated based on an average, a dot product, a variance, a standard deviation, subtraction, or addition.

[0104] In some aspects, the techniques described herein relate to a medium, wherein the distance metric is calculated based on a Euclidean distance, subtraction, addition, multiplication, a variance, or a standard deviation.

[0105] In some aspects, the techniques described herein relate to a medium, wherein the selected first entity of the plurality of first entities includes: an n-element mentee skill goal vector including n mentee skill elements, the n mentee skill elements including integer values from 1 to n; and an m-element mentee activity goal vector including m mentee activity elements, the m mentee activity elements including integer values from 1 to m; and wherein the selected second entity of the plurality of second entities includes: an n-element mentor skill goal vector including n mentor skill elements, the n mentor skill elements including values between 0 and 1; and an m-element mentor activity goal vector including m mentor activity elements, the m mentor activity elements including integer values from 1 to m.

[0106] In some aspects, the techniques described herein relate to a medium, further including electronically scheduling a meeting between each pair of matched entities.

[0107] In some aspects, the techniques described herein relate to a medium, wherein electronically scheduling a meeting includes transmitting an indication of the meeting to each entity of each pair of matched entities.

[0108] In some aspects, the techniques described herein relate to a medium, wherein electronically scheduling a meeting further includes restricting an access of a physical location based on the matching.

[0109] In some aspects, the techniques described herein relate to a medium, wherein restricting the access of the physical location includes controlling a lock.

[0110] In some aspects, the techniques described herein relate to a system for matching a plurality of first entities and a plurality of second entities, the system including: a processor; and a memory, the memory containing instructions configuring the processor to execute a method, the method including: (a) calculating, for each potential pair including a selected first entity of the plurality of first entities and a selected second entity of the plurality of second entities, a similarity metric and a distance metric; (b) ranking each potential pair relative to all other potential pairs based on the similarity metric and the distance metric for that potential pair to generate a list of ranked potential pairs; (c) nominating a top ranked pair of the list of ranked potential pairs to be matched; (d) locking any of the other potential pairs from the list of ranked potential pairs including an entity whose quota for matches has been reached to generate a list of remaining ranked potential pairs; and (e) repeating C) and D) using the list of remaining ranked potential pairs until each of the plurality of first entities or each of the plurality of second entities has been matched, (f)

[0111] In some aspects, the techniques described herein relate to a system, further including prior to A): receiving, from at least one entity of the plurality of first entities, first information associated with the at least one entity; inputting the first information associated with the at least one entity into a generative artificial intelligence (Al) model; generating, by the generative Al model, one or more responses indicating second information associated with at least one of an improved similarity metric or an improved distance metric for at least one potential pair including the at least one entity; and providing the second information to the at least one entity.

[0112] In some aspects, the techniques described herein relate to a system, wherein the generative Al model includes at least one of a large language model (LLM), a generative pretrained transformer (GPT), a generative adversarial network (GAN), a neural network, or a diffusion model.

[0113] In some aspects, the techniques described herein relate to a system, wherein inputting the first information into a generative Al model includes providing a prompt to the generative Al model.

[0114] In some aspects, the techniques described herein relate to a system, wherein the second information includes one or more suggestions for the at least one entity.

[0115] In some aspects, the techniques described herein relate to a system, wherein the one or more suggestions include a recommendation for organizing entity-specific information used during A).

[0116] In some aspects, the techniques described herein relate to a system, wherein generating, by the generative Al model, the one or more responses includes comparing the second information to a predetermined format.

[0117] In some aspects, the techniques described herein relate to a system, further including upon determining the one or more responses do not match the predetermined format, resubmitting the first information to the generative Al model.

[0118] In some aspects, the techniques described herein relate to a system, wherein: the generative Al model utilizes retrieval augmented generation (RAG) to supplement the first information with third information, the third information received from a source other than the first entity; and the generative Al model generates the second information based on the first information and the third information.

[0119] In some aspects, the techniques described herein relate to a system, wherein E) further includes at least one of transmitting a notification to an approver that a potential pair requires approval; and transmitting a notification to the first entity of each potential pair indicating that they have been matched with the second entity.

[0120] In some aspects, the techniques described herein relate to a system, further including: F) in response to the first entity not responding to the notification within a predetermined time limit, forfeiting the match between the first entity and the second entity.

[0121] In some aspects, the techniques described herein relate to a system, further including: performing A) in parallel, wherein performing A) in parallel includes reducing a time to complete A) by distributing at least a portion of A) across a plurality of processors or a plurality of cores; and E) further includes asynchronously presenting a plurality of administrators with data associated with matching the first entity and the second entity andreceiving an approval, by at least one of the administrators, of the match between the first entity and the second entity.

[0122] In some aspects, the techniques described herein relate to a system, further including: G) generating, using a large language model, at least one recommendation for a discussion topic between the first entity and the second entity based on information provided by the first entity and the second entity; and H) transmitting, to at least one of the first entity or the second entity, the at least one recommendation for a discussion topic.

[0123] In some aspects, the techniques described herein relate to a system, wherein the similarity metric is calculated based on an average, a dot product, a variance, a standard deviation, subtraction, or addition.

[0124] In some aspects, the techniques described herein relate to a system, wherein the distance metric is calculated based on a Euclidean distance, subtraction, addition, multiplication, a variance, or a standard deviation.

[0125] In some aspects, the techniques described herein relate to a system, wherein the selected first entity of the plurality of first entities includes: an n-element mentee skill goal vector including n mentee skill elements, the n mentee skill elements including integer values from 1 to n; and an m-element mentee activity goal vector including m mentee activity elements, the m mentee activity elements including integer values from 1 to m; and wherein the selected second entity of the plurality of second entities includes: an n-element mentor skill goal vector including n mentor skill elements, the n mentor skill elements including values between 0 and 1; and an m-element mentor activity goal vector including m mentor activity elements, the m mentor activity elements including integer values from 1 to m.

[0126] In some aspects, the techniques described herein relate to a system, further including electronically scheduling a meeting between each pair of matched entities.

[0127] In some aspects, the techniques described herein relate to a system, wherein electronically scheduling a meeting includes transmitting an indication of the meeting to each entity of each pair of matched entities.

[0128] In some aspects, the techniques described herein relate to a system, wherein electronically scheduling a meeting further includes restricting an access of a physical location based on the matching.

[0129] In some aspects, the techniques described herein relate to a system, wherein restricting the access of the physical location includes controlling a lock.

[0130] The present disclosure is directed towards methods and systems for enabling mentor- mentee relationships and promoting the exchange of information between employees in differing stages of their careers. Methods and systems for quantifying employee skillsets, identifying overlapping needs and interests, and mitigating unproductive matches are disclosed.

[0131] All combinations of the foregoing concepts and additional concepts discussed in greater detail below (provided such concepts are not mutually inconsistent) are part of the inventive subject matter disclosed herein. In particular, all combinations of claimed subject matter appearing at the end of this disclosure are part of the inventive subject matter disclosed herein. The terminology used herein that also may appear in any disclosure incorporated by reference should be accorded a meaning most consistent with the particular concepts disclosed herein.BRIEF DESCRIPTION OF THE DRAWINGS

[0132] The skilled artisan will understand that the drawings primarily are for illustrative purposes and are not intended to limit the scope of the inventive subject matter described herein. The drawings are not necessarily to scale; in some instances, various aspects of the inventive subject matter disclosed herein may be shown exaggerated or enlarged in the drawings to facilitate an understanding of different features. In the drawings, like reference characters generally refer to like features (e.g., functionally similar and / or structurally similar elements).

[0133] FIG. 1 A illustrates a pool of first entities and a pool of second entities.

[0134] FIG. IB illustrates a data structure representing potential pairs of first entities and second entities.

[0135] FIG. 1C illustrates a data structure including data representing parameters that may be used to identify potential matches between first entities and second entities.

[0136] FIG. ID illustrates multiple first entity perspective ranking groups including a selected first entity perspective ranking group for a selected first entity with respect to one or more second entities.

[0137] FIG. IE illustrates blocks of a method for second entity perspective pair analysis analogous to those for the first entity perspective for creating a combined first and second entity perspective rank.

[0138] FIG. IF illustrates a first entity perspective pair array and a second entity perspective pair array and the accompanying perspective ranks for each perspective and pair.

[0139] FIG. 1G further illustrates the block of averaging first and second perspective ranks to determine a global ranking metric.

[0140] FIG. 1H illustrates a process of selecting entity pairs from a list organized by global ranking metric.

[0141] FIG. II illustrates a continued process of nominating a subsequent entity pair to be matched based on a previous nomination block.

[0142] FIG. 1 J illustrates a final nomination block and exemplary matched pairs.

[0143] FIG. 2 illustrates an exemplary system for performing entity matching in accordance with the disclosure.

[0144] FIGS. 3A and 3B illustrate a method for matching in accordance with the present technology.

[0145] FIG. 4 illustrates a method for matching in accordance with the present technology.DETAILED DESCRIPTION

[0146] FIGS. 1 A-1I illustrate aspects of a computer-implemented method 100 for mentor- mentee matching that may be carried out by a computer system in accordance with the present disclosure. Any or all steps, blocks, actions, or processes recited herein may be performed using one or more processors, computer systems, cloud computing systems, central processing units (CPUs), storage units, field-programmable gate arrays (FPGAs), integrated circuits (ICs), graphics processing units (GPUs), or similar computing systems such as those illustrated by FIG. 2. Method steps, blocks, actions, and / or processes may be stored on one or more non-transitory computer-readable media as instructions for configuring one or more processors to execute the respective method steps, blocks, actions, and / or processes. The present technology is broadly directed to systems and methods for generating final accepted matches (FAMs) of pairs of entities by first filtering a first group of entitiesand a second group of entities to find eligible and compatible potential pairs, and then ranking those potential pairs as a function of how closely the priorities or matching characteristics of the entities of the potential pairs align with each other. Subsets of the potential pairs may then be nominated for matching. The nominated pairs may then be accepted or rejected, e.g., accepted by one or more of the entities in the pairs and / or one or more administrators.

[0147] These entities (e.g., mentees and / or mentors) may input first information including employment and demographic information, work priorities, personal development priorities, personal interests, and personal characteristics into an application interface such as a website. The first information may then be used to perform matching and pair generation. In an aspect, the present systems and methods may utilize generative artificial intelligence (Al) models including as one or more large language models (LLMs), generative pre-trained transformers (GPTs), generative adversarial networks (GANs), neural networks, diffusion models, and the like, to utilize the first information to provide suggestions for how mentee applicants should provide information to optimize a mentor match for that mentee. As used herein, references to “a generative Al model” may be used to refer to the model itself, or to a computer implementation operating the model (e.g., a processor utilizing the model to perform calculations and / or generate outputs).

[0148] The first information provided by a mentee applicant may include a prompt (which may be in a natural language format) to a generative Al model. An example prompt may be “If I want to match with a mentor in the legal department, how should I indicate that?”

[0149] In response, a prompt provided to a generative Al model may cause the generative Al model to provide second information. For example, the generative Al model may provide a suggestion output to the mentee applicant in a suitable form such as a nested list of JavaScript Object Notation (JSON) data or other suitable structured data format. A nested list of JSON data may include one or more key: value pairs enclosed within brackets, such as: { “Include” : { “username” : “Martin Smith”, “work_information” : {“department” : “Legal”, “supervisor” : “Morgan Freeman”, “years_experience” : “16”} }, “Exclude” : {“work_city” : “Boston” }, “Unrelated” : {“active_directory_group” : “Life Sciences”, “parent_organization” : “Regeneron”, “work_country” : “U.S.A.”} }. The suggestion output may be compared to the structured data format (e.g., if a suggestion output should match JSON data, a processor may determine if the suggestion output consists of key:value pairswithin curly brackets), and the suggestion output presented to the mentee applicant or user who presented the prompt and / or requested the output.

[0150] However, due to the nature of generative Al suggestion output (which may be based on dynamic probabilities that are a function of training data, model weights, prompts, and other factors), the suggestion output provided by a generative Al model may not be in the appropriate predetermined format (e.g., may not have the form of a nested list of JSON data). In such an instance, the present systems and methods may resubmit the prompt to the generative Al model and check if the suggestion output received from the generative Al model matches the appropriate format. If the suggestion output received still does not match the appropriate format, the prompt may be input into the generative Al model a third or subsequent time to attempt to receive a response in the required format. In an embodiment, a predetermined number of prompts may be input into the generative Al model. If no suggestion output having the appropriate format is received, a notification may be provided to the person or entity requesting the input that their request was unsuccessful.

[0151] Determining if a suggestion output matches an appropriate format may include evaluating whether the suggestion output from the generative Al model starts or contains one or more specific characters, numbers, letters, symbols, etc. For example, a processor (such as processor 210 shown in FIG. 2) may evaluate whether the suggestion output starts with a curly bracket “{“ corresponding to JSON-formatted data. If the suggestion output does not start with a curly bracket (or does not contain a curly bracket), the processor may determine that the suggestion output provided by the generative Al model is not in JSON format and therefore does not include the information required by the person or entity who placed the request via the prompt.

[0152] A suitable suggestion output may include a list of compatibility categories that may be used for determining inclusion criteria, exclusion criteria, and irrelevant criteria. For example, if an entity such as a mentee applicant inputs a prompt asking “If I want to match with a mentor in the legal department, how should I indicate that?” a suitable suggestion output may include inclusion criteria, exclusion criteria, and irrelevant criteria. Example inclusion criteria may include job profile, location, and / or work department; example exclusion criteria may include a potential mentor’s username or a username of a potential mentor’s supervisor; and example criteria indicated as irrelevant and may include categories such as active directory group, company, parent organization, work city, work country, cost center identification number and name, job familyjob family group, organization, etc.

[0153] A generative Al model may be or utilize a fine-tuned LLM (e.g., an LLM trained on both a generalized training dataset as well as a more specific training dataset). Additionally or alternatively, a generative Al model may utilize retrieval augmented generation (RAG), which may be able to consider thousands of choices that are selectable by a system user such as a mentee applicant. RAG may enable a generative Al model to respond with greater detail to a prompt, including more detailed instructions on how to define compatibilities within applications as compared to a non-RAG generative Al model.

[0154] A generative Al model used in accordance with the present technology may utilize RAG to receive third information from a source other than the first entity and supplement first information received from the first entity. The generative Al model may utilize a combination of at least a portion of the first information and at least a portion of the third information to generate second information. For example, the first entity may be a mentee applicant who submits first information such as demographic information specifying their name, department, and requested mentor department. The first information may further include a natural language prompt such as “I would like to be matched with a mentor in the R&D department with at least 5 more years of experience than I have.” The generative Al model may then retrieve third information from a source other than the first entity (e.g., a human resources database) indicating a length of tenure that the mentee applicant has. The third information may further include application options that the mentee applicant may specify, such as where and how a mentee applicant may specify a preferred length of tenure of a mentor.

[0155] FIG. 1 A illustrates a pool of first entities 101 and a pool of second entities 102. The first entities may be mentors and the second entities may be mentees. One or more of the mentors may be matched with one or more of the mentees to create a mentor-mentee pair and enable a mentorship relationship to benefit a mentee and spur the mentee’ s career growth. To determine this matching, a group of mentors and mentees may be compared based on one or more eligibility parameters and compatibility parameters. For example, each mentee of a group of mentees may be eligible to be matched with a mentor from a group of mentors based on eligibility and compatibility parameters (including ranges of such parameters) including a role, a job title, a division or group within an organization, compensation grade (e.g., a range of compensation grades), years of experience (e.g., a range of years of experience), supervisor chain, desired skill set or improvement to a skill, name, username, reporting structure hierarchy, department, parent organizationjob family group, job family, costcenter, employee type, reporting line, gender, age, entity personal preference, mentee- or administrator-specified inclusion or exclusion criteria, or any suitable identification and / or selection criteria. Eligibility may be determined automatically by one or more systems in accordance with the present technology; for example, a processor may implement eligibility parameters to select a group of eligible mentee candidates and a group of eligible mentor candidates.

[0156] Method 100 may include identifying a group of first entities (such as a group of mentees 120 including mentees A, B, and C) from the overall pool of first entities based on one or more parameters such as the eligibility parameters recited above. Method 100 may further include identifying a corresponding group of second entities (such as a group of eligible mentors 110 including mentors 1, 2, 3, 4, 5, and 6) from the overall pool of second entities based on one or more parameters such as the eligibility parameters recited above.

[0157] FIG. IB illustrates a data structure representing potential pairs of first entities and second entities, for example mentee-mentor pairs. In an embodiment, the data structure may be an array (e.g., a mentee perspective pair array 130), a list, a dictionary, a key-value pair structure, a linked list, a tuple, a tree, or any suitable data structure. Each entity of the groups of first and second entities may be subject to an eligibility filter to first determine if each respective entity meets one or more requirements to be considered for initial pairing. Then, each entity of the group of first entities may be compared to each entity of the group of second entities (and vice versa) to ensure that a potential match between the respective entities meets both eligibility criteria and compatibility criteria (criteria may refer to threshold conditions or ranges for parameters, e.g. a parameter may be years of experience and a criterion may be more than 10 years, or between 10 and 20 years, of experience).

[0158] For example, mentor 1 meets eligibility criteria and compatibility criteria to be considered as a pair with mentee A; such pairs may be represented as A / l in this instance, but does not meet eligibility criteria and compatibility criteria to be considered as a pair with mentee B. Therefore, pair A / l is available to be considered and is included in mentee perspective pair array 130 (and mentor perspective pair array 140 in FIG. IE), but pair B / l is not available to be considered for matching and is therefore absent from mentee perspective pair array 130 (and mentor perspective pair array 140 in FIG. IE).

[0159] Method 100 may include generating a pair array from the group of eligible mentors 110 and group of mentees 120, each pair array organized by “perspective,” e.g., listing eachof the mentees in order in a first column and listing each possible mentor for that mentee in a second column to create a mentee perspective pair array 130. For example, a generated pair array may list mentee A in a first column and list each eligible and compatible mentor (e.g., selected from mentors 1-6) in a second column next to mentee A. Likewise, a mentor perspective pair array 140 may be created in a similar way, listing each of the mentors in order in a first column and listing each eligible and compatible mentee for that mentor in a second column. An example of a mentor perspective pair array 140 is illustrated in FIG. IE.

[0160] Method 100 may include identifying a selected entity from the group of first entities. For example, method 100 may include identifying a selected mentee from the group of mentees 120 (which may correspond to the mentees in mentee perspective pair array 130). Identifying a selected mentee may occur once a plurality of mentees and mentors has been filtered and arranged into a mentee perspective pair array 130. Each mentee may have one or more eligible and compatible mentors, selected for eligibility from the pool of mentors 101 and further selected for compatibility from the group of eligible mentors 110), and the one or more eligible and compatible mentors may be selected from the pool of mentors by a processor based on eligibility and compatibility criteria, predetermined parameters, conditions imposed by an administrator, or any suitable selection rationale or mechanism.

[0161] FIG. 1C illustrates a data structure including skill and activity parameters that may be used to identify potential matches between first entities and second entities. In an embodiment, the data structure may include firs entity data and second entity data respectively identifying one or more first entities and one or more second entities. Each first entity data and second entity data may include respective entity parameter vectors representing one or more aspects of each entity. Each respective parameter vector representing each entity may be compared with a corresponding parameter vector for a different entity to determine a similarity metric, a distance metric (e.g., a value representing how much two entities differ in one or more aspects), or any suitable comparison value. Comparing parameter vectors may include calculating a dot product, a Euclidean distance, an addition of two vectors, a subtraction of two vectors, a cross product, a magnitude, an average, a median, a mode, a distribution, a standard deviation, a variance, a proportion, a correlation, a regression coefficient, an analysis of variance (ANOVA) calculation, or any suitable calculation.

[0162] For example, the data structure may be mentee perspective pair array 130 and include respective entity parameter vectors corresponding to a first entity or group of firstentities (such as group of mentees 120) and a second entity or group of second entities (such as group of eligible mentors 110), as well as calculations performed with respect to a method 100 in accordance with the disclosure. Each respective entity parameter vector may include one or more elements.

[0163] In an embodiment, data for each mentor and each mentee of the mentee perspective pair array 130 may include an n-element skill goal vector 126. Each mentee skill goal vector 126 may include one or more elements representing desired skill sets that the mentee wishes to improve, such as technical writing, public speaking, networking, data analysis, and the like. Each element of mentee skill goal vector 126 may be a number including a decimal, an integer, a fraction, a rating, a semantic descriptor, a scale value, a rational number, an irrational number, a hexadecimal number, a decimal number, a binary number, and may be ranked based on value, order of entry, in ascending order, in descending order, in exclusive order, in overlapping order, or any suitable order.

[0164] Each element of a mentee skill goal vector 126 may be an integer from 1 to n indicating a mentee’ s ranking of importance of a particular skill set that the mentee wishes to improve, with n indicating the greatest desire to improve and 1 indicating the least desire to improve. For example, mentee skill goal vector 126 may have six elements representing a mentee’ s desire to improve the categories of technical writing, public speaking, networking, data analysis, email communication, and project management, which the mentee may rank as [3, 2, 5, 4, 1, 6], This may indicate that a mentee is (relatively) least interested in improving their email communication skill set and is most interested in improving their project management skill set.

[0165] Each mentor skill goal vector 116 may include one or more elements representing an existing skill set and / or comfort level mentoring a mentee in skill set corresponding to the skill sets included in the mentee skill goal vector 126. Each element of a mentor skill goal vector 116 may be 0, 0.5, or 1, and multiple elements may have the same value. 0 may indicate that a mentor is not comfortable mentoring a particular skill set, 0.5 may indicate that a mentor is marginally comfortable mentoring a particular skill set, and 1 may indicate that a mentor is fully comfortable mentoring a particular skill set. For example, a mentor skill goal vector 116 may have six elements representing a mentor’s willingness or ability to provide mentorship in the same categories of technical writing, public speaking, networking, data analysis, email communication, and project management, which the mentor may rank as [0.5, 1, 0.5, 0, 0, 1], This may indicate that a mentor is fully confident mentoring public speakingand project management, marginally comfortable mentoring technical writing and networking, and not comfortable mentoring data analysis and email communication.

[0166] Method 100 may include calculating, for each pair represented by the data structure, a comparison value indicating a similarity between the two entities in that pair, for example a similarity metric. The comparison value may be calculated based on respective parameter vectors for the two entities being compared. The comparison value may be calculated from the perspective of either entity in the pair. For example, method 100 may include calculating, for each potential mentee-mentor pair 132 of the mentee perspective pair array 130 including the selected mentee, a dot product of mentee skill goal vector 126 for the selected mentee and mentor skill goal vector 116 for the mentor of that pair to generate a similarity metric such as skill similarity 136 for that pair. The skill similarity 136 for a pair may indicate a degree to which a skill set that a mentee desires to improve overlaps with a mentor’s ability to help the mentee improve that skill set, with a higher score indicating a more desirable mentor-mentee match. A skill similarity is maximized when a mentor feels fully confident that they can mentor each skill set indicated by the mentor and mentee n-element skill goal vectors (and therefore rates each mentor existing skill set “1”), but may also indicate a strong overlap when a mentor feels fully confident about mentoring the same skill sets that a mentee is prioritizing.

[0167] A similarity metric may be calculated using any suitable comparison. For example, a similarity metric may be calculated using a dot product, a Euclidean distance, an addition of two vectors, a subtraction of two vectors, a cross product, a magnitude, an average, a median, a mode, a distribution, a standard deviation, a variance, a proportion, a correlation, a regression coefficient, an analysis of variance (ANOVA) calculation, or any suitable calculation.

[0168] For example, for a generalized 6-element skill goal vector representing 6 skill sets, the maximum skill similarity for the pair would be the dot product of [1, 2, 3, 4, 5, 6] and [1, 1, 1, 1, 1, 1], which equals 21. However, even if a mentor only feels fully confident about mentoring half of the skill sets (e.g., has a mentor n-element skill goal vector of [0, 0, 0, 1, 1, 1]), if those mentor skill sets with full confidence correspond to a mentee’ s most highly desired skill sets (e.g., the skill sets rated 4, 5, and 6), a skill similarity for the pair will be 15, which is 71% of the maximum score despite a mentor skill goal vector having only 50% of the maximum element values.

[0169] FIG. 1C illustrates a mentee skill goal vector 126 of [3, 2, 5, 4, 1, 6] for mentee A and a mentor skill goal vector 116 of [0.5, 1, 0.5, 0, 0 1] for mentor 1. The skill similarity for the A / l pair is calculated as the dot product of the mentee skill goal vector 126 and the mentor skill goal vector 116, which has a value of 12.

[0170] Method 100 may further include calculating, for each pair represented in a data structure, a distance metric representing a difference between entities in that pair. For example, method 100 may further include calculating, for each pair of the mentee perspective pair array 130 including the selected mentee, a Euclidean distance of an m-element mentee activity goal vector 128 for the selected mentee and an m-element mentor activity goal vector 118 for the mentor of that pair to generate an activity distance 138 for that pair. A distance metric may be calculated using any suitable comparison. For example, a distance metric may be calculated using a dot product, a Euclidean distance, an addition of two vectors, a subtraction of two vectors, a cross product, a magnitude, an average, a median, a mode, a distribution, a standard deviation, a variance, a proportion, a correlation, a regression coefficient, an analysis of variance (ANOVA) calculation, or any suitable calculation.

[0171] Each mentor activity goal vector 118 may include m mentor activity elements. The m mentor activity elements may include integer values from 1 to m. Each mentee activity goal vector 128 may include m mentee activity elements. The m mentee activity elements may include integer values from 1 to m. An activity distance 138 may indicate a difference between a mentee’ s interest in one or more activities to develop a desired skill set and a mentor’s comfort with participating in or coaching the one or more activities. A lower score indicates a more desirable alignment between a mentor and mentee. An activity distance 138 may be calculated(the Euclidean distance for two vectors), where ariis the ithelement of a mentor activity goal vector 118 and aeiis the ithelement of a mentee activity goal vector 128. An activity distance 138 with the minimum value of 0 indicates a maximum degree of alignment between a mentor and mentee.

[0172] For example, FIG. 1C illustrates a mentee activity goal vector 128 (having 6 elements) including elements [6, 5, 1, 3, 2, 4] and a mentor activity goal vector 118 (having 6 elements) including elements [2, 6, 5, 3, 4, 1], These elements may correspond to activities that may be performed to improve skill sets corresponding to the mentee skill goal vector 126; however, an activity goal vector may have multiple activities corresponding to a single skill set represented by a skill goal vector or vice versa (a single activity may apply tomultiple skill sets). For example, the activities of the mentee and mentor activity goal vector may include writing a white paper, giving a talk at a conference, attending at least three networking events, analyzing drug trial results in a particular programming language, writing at least ten practice emails, and shadowing a mentor during the management of a project. An activity distance 138 for a potential mentee-mentor pair 132 may indicate that a mentor and mentee have overlapping activity interests and skills. Using the above mentor and mentee activity goal vectors, an activity distance 138 for mentee A and mentor 1 in FIG. 1C is calculated as 6.78.

[0173] FIG. ID illustrates the results of further calculations for each pair of the mentee perspective pair array 130 corresponding to the skill similarity and activity distance calculations of FIG. 1C as well as ranking and averaging calculations to generate a mentee perspective rank of potential mentors.

[0174] Method 100 may further include calculating a skill similarity 136 and activity distance 138 for each eligible mentor for a selected mentee. For example, mentee A may potentially match with mentors 1, 2, 4, and 6. The skill similarities and activity distances for mentee A paired alternatively with mentors 2, 3, and 4 may be calculated and determined to be the following:For mentee A and mentor 2, a skill similarity of 8.5 and an activity distance of 7.24.For mentee A and mentor 4, a skill similarity of 9 and an activity distance of 8.14.For mentee A and mentor 6, a skill similarity of 9.5 and an activity distance of 4.25.

[0175] Method 100 may further include generating, for each pair of entities, a ranking pair, the ranking pair comprising a ranking of the similarity metric and the distance metric for that pair. For example, method 100 may further include generating, for each pair 132 of the mentee perspective pair array 130 including the selected mentee, a ranking pair 135, the ranking pair 135 comprising a ranking of the skill similarity 136 and the activity distance 138 for that pair of the mentee perspective pair array 130. Method 100 may further include placing the ranking pair into a mentee perspective ranking group 134a-c. Each of the skill similarities and activity distances for a selected mentee and an eligible mentor for the selected mentee may be ranked relative to the skill similarities and activity distances for the other eligible mentors for that mentee. For example, the skill similarities and activity distances of mentee A and mentee A’s eligible mentor pairs may be ranked in order in mentee perspective ranking group 134a, with highest skill similarity being ranked 1 and the lowest skill similaritybeing ranked 4, and the lowest activity distance being ranked 1 and the highest activity distance being ranked 4.

[0176] FIG. ID illustrates multiple mentee perspective ranking groups 134a-c including a mentee perspective ranking group 134a for mentee A with respect to mentors 1, 2, 4, and 6. Each ranking group may include a rank of each skill similarity 136 for the eligible mentors 1, 2, 4, and 6. A rank of 1 may indicate the best match for a selected mentee, with a higher (better) skill similarity 136 corresponding to a lower (better) rank. From mentee A’s perspective, mentor 1 has the best skill similarity 136 with a value of 12, followed by the skill similarity 136 for mentor 6 with a value of 9.5, followed by the skill similarity for mentor 4 with a value of 9, and lastly the skill similarity for mentor 2 with a value of 8.5. Accordingly, the skill similarity 136 for mentor 1 would be ranked 1, the skill similarity for mentor 6 would be ranked 2, the skill similarity for mentor 4 would be ranked 3, and the skill similarity for mentor 2 would be ranked 4.

[0177] The activity distances 138 for each mentee and eligible mentor pair may be similarly ranked. The mentee perspective ranking groups 134a-c may include a rank of each activity distance for the mentors with which they are eligible and compatible. A rank of 1 may indicate the best match for a selected mentee, with a lower (better) activity distance 138 corresponding to a lower (better) rank. From mentee A’s perspective, mentor 6 has the best activity distance 138 with a value of 4.25, followed by the activity distance 138 for mentor 1 with a value of 6.78, followed by the activity distance 138 for mentor 2 with 7.24, and lastly the activity distance 138 for mentor 4 with 8.14. Accordingly, the activity distance 138 for mentor 6 would be ranked 1, the activity distance 138 for mentor 1 would be ranked 2, the activity distance 138 for mentor 2 would be ranked 3, and the activity distance 138 for mentor 4 would be ranked 4. From mentee A’s perspective, the skill similarity 136 and activity distance 138 for mentors 1, 2, 4, and 6 would therefore be:Mentor 1: 6.78-Mentor 2: 7.24Mentor 4: 8.14Mentor 6:4.25-

[0178] Method 100 may further include calculating a combination metric for each ranking pair within a ranking group. A combination metric may be calculated by applying a mathematical operation to the elements in each ranking pair. For example, an average, median, mode, variance, standard deviation, proportion, percentage, addition, subtraction, division, or multiplication of the elements in each ranking pair may be calculated. Forinstance, method 100 may further include averaging, for each pair of mentee perspective ranking groups 134a-c, the two rankings of skill similarity 136 and activity distance 138 to calculate a merged mentee perspective rank. For mentee A, the mentee perspective ranking group 134a may be:Mentor 1: 1 2-Mentor 2: 4 3Mentor 4: 3 4Mentor 6: -2 1-Each row of the mentee perspective ranking group 134a may be averaged to provide a merged mentee perspective rank 137a. The merged mentee perspective ranking 137a for mentee A is accordingly:Mentor 1: 1.5-Mentor 2: 3.5Mentor 4: 3.5Mentor 6: -1.5-

[0179] Method 100 may further include repeating the above calculating, generating, and averaging steps for each additional mentee of the plurality of mentees. For example, the above calculating, generating, and averaging may be repeated for mentee B and C with respect to each eligible and compatible mentor for those mentees to produce mentee perspective ranks for mentees B and C analogous to those for mentee A as illustrated in FIG. ID. Repeating the calculation steps illustrated in FIG. 1C for mentee B, the skill similarity 136 and activity distance 138 for mentors 3, 4, and 6 from mentee B’s perspective may be:Mentor 3: [4.5 7.91Mentor 4: 14 3.24Mentor 6: 10 5.25-For mentee B, the mentee perspective ranking group 134b is therefore:Mentor 3: [3 3Mentor 4: 1 1Mentor 6: 1-2 2- and the merged mentee perspective rank 137b for mentee B is accordingly:Mentor 3: 3Mentor 4: 1Mentor 6: 1-2-1Repeating the calculation steps illustrated in FIG. 1C for mentee C, the skill similarity 136 and activity distance 138 for mentors 1-3 and 5-6 from mentee C’s perspective may be:Mentor 1: - 2.5 4.81-Mentor 2: 12.5 3.32Mentor 3: 18 7.98Mentor 5: 8.5 3.38Mentor 6: - 6.5 8.22 -For mentee C, the mentee perspective ranking group 134c would therefore be:Mentor 1: 3 r5Mentor 2: 2 1Mentor 3: 1 4Mentor 5: 3 2Mentor 6: -4 5-1 and the merged mentee perspective rank 137c for mentee C is accordingly:Mentor 1: r4]Mentor 2: 1.5Mentor 3: 2.5Mentor 5: 2.5Mentor 6: U.5-1

[0180] FIG. IE illustrates steps of method 100 for mentor perspective pair analysis analogous to those for the mentee for creating a merged mentor perspective rank.

[0181] Method 100 may include generating a mentor perspective pair array 140, the mentor perspective pair array 140 comprising the plurality of eligible and compatible mentors of the plurality of mentors and the plurality of mentees represented as pairs. The mentor perspective pair array 140 may contain each of the same pairs as the mentee perspective pair array 130, but arranged in order by mentor instead of mentee. Additionally or alternatively, the mentor perspective pair array 140 may be the same as or similar to the mentee perspective pair array 130. In an embodiment, information represented by the mentor perspective pair array 140 may be stored in the same format as the information represented by the mentee perspective pair array 130.

[0182] In an embodiment, a processor (such as processor 210 shown in FIG. 2) may instantiate data structures such as mentee perspective pair array 130 and mentor perspective pair array 140 by accessing the same data in a different order based on the pair array to be instantiated. For example, mentor and mentee data may be stored on a storage unit (such as a flash drive, a solid state drive (SSD), a server, a distributed cloud storage system, or similar storage system such as storage unit 230 shown in FIG. 2) and a processor may instantiate one or more data structures such as mentee perspective pair array 130 and / or mentor perspective pair array 140 on a memory (such as random access memory (RAM), embodied as memory 220 in FIG. 2) as needed.

[0183] Method 100 may include identifying a selected mentor from the plurality of eligible and compatible mentors 110. Identifying a selected mentor may occur once a plurality of mentors and mentees has been determined and arranged into a mentor perspective pair array140. Each mentor may have one or more eligible and compatible mentees 120, selected from the pool of mentees 102, and the one or more eligible and compatible mentees 120 may be selected from the pool of mentees 102 by a processor based on one or more eligibility criteria and one or more compatibility criteria, predetermined parameters, conditions imposed by an administrator, or any suitable selection rationale or mechanism.

[0184] Method 100 may include calculating, for each pair of the mentor perspective pair array 140 including the selected mentor, a dot product of an n-element mentor skill goal vector 116 for the selected mentor and an n-element mentee skill goal vector 126 for the mentee of that pair to generate a skill similarity 136 for that pair. The calculation of skill similarity 136 may be the same as or similar to the analogous calculation block from the mentee perspective illustrated in FIG. 1C. In an embodiment, the skill similarity 136 and activity distance 138 may be the same value for both the mentor-mentee perspective as well as the mentee-mentor perspective for a given pair. For example, mentee A and mentor 1 may have a skill similarity 136 of 12 and an activity distance 138 of 6.78 regardless from which perspective the pair is analyzed.

[0185] In an embodiment, the block of calculating, for each pair of the mentor perspective pair array 140 including the selected mentor, a dot product of a mentor skill goal vector 116 for the selected mentor and a mentee skill goal vector 126 for the mentee of that pair to generate a skill similarity 136 for that pair may be omitted or replaced by retrieving the skill similarity 136 calculated for that pair from the mentee perspective using the mentee perspective pair array 130.

[0186] Method 100 may include calculating, for each pair of the mentor perspective pair array 140 including the selected mentor, a Euclidean distance of a mentor activity goal vector 118 for the selected mentor and a mentee activity goal vector 128 for the mentee of that pair to generate an activity distance 138 for that pair. The calculation of activity distance 138 may be the same as or similar to the analogous calculation block from the mentee perspective illustrated in FIG. 1C. In an embodiment, the skill similarity 136 and activity distance 138 may be the same value for both the mentor-mentee perspective as well as the mentee-mentor perspective for a given pair. For example, mentee A and mentor 1 may have a skill similarity 136 of 12 and an activity distance 138 of 6.78 regardless from which perspective the pair is analyzed.

[0187] In an embodiment, the block of calculating, for each pair of the mentor perspective pair array 140 including the selected mentor, a Euclidean distance of a mentor activity goal vector 118 for the selected mentor and a mentee activity goal vector 128 for the mentee of that pair to generate an activity distance 138 for that pair may be omitted or replaced by retrieving the activity distance 138 calculated for that pair from the mentee perspective using the mentee perspective pair array 130.

[0188] Method 100 may include generating, for each pair of the mentor perspective pair array 140 including the selected mentor, a ranking pair, the ranking pair comprising a ranking of the skill similarity 136 and the activity distance 138 for that pair of the mentor perspective pair array 140, and placing the ranking pair into a mentor perspective ranking group (such as mentor perspective ranking groups 144a-f). This block may be analogous to the block of generating the ranking pair from the mentee perspective illustrated in FIG. ID.

[0189] Method 100 may further include calculating a combination metric for each ranking pair within a mentor perspective ranking group. A combination metric may be calculated by applying a mathematical operation to the elements in each ranking pair. For example, an average, median, mode, variance, standard deviation, proportion, percentage, addition, subtraction, division, or multiplication of the elements in each ranking pair may be calculated.

[0190] For example, method 100 may further include averaging, for each pair of rankings in mentor perspective ranking group 144a-f, the rankings of that pair to determine merged mentor perspective ranks 147a-f. FIG. IE illustrates further calculations for each pair of the mentor perspective pair array 140 corresponding to the skill similarity and activity distance calculations of FIG. 1C as well as ranking and averaging calculations to generate a mentor perspective rank of potential mentees.

[0191] The block of averaging, for each pair of the mentor perspective ranking group, the rankings of that pair to determine a merged mentor perspective rank illustrated in FIG. IE may be the same as or analogous to the corresponding block for each pair of the mentee perspective ranking group illustrated in FIG. ID. Each block described above with respect to the mentor perspective pair array 140 (including the calculating, generating, and averaging blocks) may be repeated for each additional mentor of the plurality of eligible and compatible mentors.

[0192] Method 100 may further include calculating a global combination metric for each perspective for a given pair of entities. A combination metric may be calculated by applying a mathematical operation to the elements in each ranking pair. For example, an average, median, mode, variance, standard deviation, proportion, percentage, addition, subtraction, division, or multiplication of the elements in each ranking pair may be calculated. For example, method 100 may further include a block of averaging, for each possible mentor- mentee pair from the plurality of mentees and the plurality of eligible and compatible mentors, the mentor perspective rank and the mentee perspective rank to determine a global ranking metric 150, illustrated in FIG. 1G. FIG. IF illustrates the mentee perspective pair array 130 and the mentor perspective pair array 140 and the accompanying perspective ranks for each perspective and pair.

[0193] FIG. 1G further illustrates the block of averaging, for each possible mentor-mentee pair from the plurality of mentees and the plurality of eligible and compatible mentors, the mentor perspective rank and the mentee perspective rank to determine a global ranking metric 150. For example, the ranking of mentor 4 from the perspective of mentee B (with a value of 1) is compared to the ranking of mentee B from the perspective of mentor 4 (also with a value of 1). The two values may be averaged to determine a combined ranking. The combined ranking for each pair may be ordered to produce a global ranking metric 150. In this case, the global ranking metric 150 for mentee B and mentor 4 is 1 (the best rank average). This calculation is repeated for each possible mentor-mentee pair.

[0194] Once the global ranking metric 150 has been calculated for each mentor-mentee pair, the pairs in the metric 150 may be reordered from lowest to highest. FIG. 1G illustrates each of the pairs ordered by score, with pair B / 4 ranked first with a global ranking metric value of 1, pair A / l and C / 2 tied for second with a global ranking metric value of 2, and pair A / 6 ranked third with a global ranking metric value of 3, and so on. In cases where two or more pairs are tied, compatibility preferences associated with mentees, mentors, and other entities, data provided by nominated entities, and / or manual administrator inputs may be used to break ties.

[0195] Method 100 may further include generating one or more final accepted matches (FAMs) by matching one or more first nominated entities of the first entities with one or more second nominated entities of the second entities based on the global ranking metric 150. For example, method 100 may further include matching a mentee of the plurality of mentees with a mentor of the plurality of eligible and compatible mentors based on the global rankingmetric 150. Each match may be initiated by nominating a pair to be matched and then temporarily locking any additional pairs containing a mentee and / or mentor whose quota has been reached.

[0196] Some or all portions of blocks of method 100 may happen in parallel or asynchronously for each potential pair. For example, a skill similarity and activity distance calculation block illustrated in FIG. 1C may be performed in parallel for each potential mentee-mentor pair 132 in mentee perspective pair array 130. Calculations of skill similarity 136 and activity distance 138 for each respective pair may accordingly be assigned to individual cores in a processor, to individual processors (e.g., in a distributed computing system), or distributed in any suitable way such that a plurality of the calculations may be performed substantially simultaneously. Both parallelization and asynchronous processing may improve time cost, speed, and computational efficiency by deserializing some or all portions of blocks of method 100. For example, processing a plurality of mentee-mentor pair skill similarity and activity distance blocks in parallel may reduce a computation time for all mentee-mentor pairs by O(N) where N is a number of individual cores or processor threads that can simultaneously perform the skill similarity and activity distance blocks. In an embodiment, a processor may distribute parallelizable tasks across a suitable number of compute cores (e.g., 15 or more compute cores), which may reduce overall block calculation time by lOx or more. Additional complexities of parallelization and asynchronization due to task scheduling, memory allocation, and added communication times are fully offset by a reduction in computing time may drastically improve system performance and user experience including through faster matching and more efficient resource utilization.

[0197] Further, some or all blocks of method 100 may be performed asynchronously while the new information is being input into a system (e.g., system 200 as illustrated in FIG. 2). For example, mentee perspective ranking groups for one or more existing mentee applicants may be calculated while the system transmits and receives information for an application from a new mentee applicant. Once the new mentee applicant has submitted their application, the relevant blocks of method 100 may be calculated to incorporate the new mentee applicant into mentee perspective ranking groups, mentor perspective ranking groups, global ranking metrics, and the like.

[0198] FIG. 1H illustrates a process of selecting mentor-mentee pairs from a list organized by global ranking metric 150. In a first block, the highest-ranked mentor-mentee pair may be nominated to be matched, which is mentee B and mentor 4. Each mentor and mentee mayhave a distinct, respective quota for a number of entities with which they will be nominated and matched. For example, a mentee may have a quota of one mentor, meaning they will only be nominated and matched with one mentor and then their candidacy for additional nominations will be temporarily locked (removed from consideration) unless the initially- nominated match has been denied. A nominated pair may be denied by an administrator, by a software program (for a variety of reasons, such as automatic deadline-based forfeiture), or by either of the mentor or mentee of the pair. If a nominated pair is denied, that individual pair will remain locked permanently, but the nomination count for each of the members of that denied pair will be reduced by 1, resulting in the reversal of the lock for all other pairs containing either of the members of that denied pair, but only for those other pairs that do not also contain other members whose quotas have been reached.

[0199] A nominated pair must be accepted by the mentee of the pair. In an embodiment, a nominated pair must additionally be accepted by an administrator, which may include one or more of a nominator, a reviewer, an approver, or the like. A mentor may have more than one mentee, such as two, three, or more mentees. In an aspect, a notification may be automatically generated based on a suitable step in method 100; for example, if a nominated pair requires approval from an administrator, a mentee, a human resources business partner (HRBP), etc. HRBPs may be reviewers, nominators, or have any other suitable role in a process of matching mentees and mentors.

[0200] Nomination of pairs of entities (e.g., mentees and mentors) for matching may be at least partially automated. An exemplary nomination process may be automatically triggered a predetermined period of time after an application period (e.g., a time period in which first entities and second entities may submit information that may be used to determine matching pairs of first and second entities). This automation and associated tie-breaking criteria may ensure improved efficiency, increased fairness to paired entities, reduction in manual intervention, and reduction in subjective and inconsistently applied match criteria.

[0201] For example, an exemplary nomination process may be automatically triggered one day after an application period closes. Additionally or alternatively, an automated nomination process may occur in response to an administrator-based trigger (e.g., in response to an approver triggering the process), or in response to a manual input through, e.g., an application programming interface (API) endpoint activated by a suitable mechanism such as a button in a user interface. Such a user interface provides flexibility and control for human users such as HRBPs.

[0202] A tie-breaking global rank (equivalently, global rank with ties broken [GRTB]) calculation may be performed on rankings of pairs of entities. The tie-breaking global rank calculation may be based on a first entity’s (a mentee’s) relative rank, a second entity’s (a mentor’s) relative rank, the first entity's tenure or association with an organization, the second entity's tenure or association with an organization, alphabetical order of first entity username, and alphabetical order of second entity username.

[0203] Once a GRTB is calculated, the entity pair with the highest GRTB rank may be nominated automatically. For each nominated entity pair, LLM-based discussion topics may be generated and database caching of discussion topic recommendations may be performed. These discussion topics may be tailored to each entity of each nominated pair, and further may be generated based on information associated with each entity of each nominated pair as well as how the information for the first entity and information for the second entity interrelates. A first set of discussion topic recommendations may be tailored and / or generated for the first entity, while a second set of discussion topic recommendations may be tailored and / or generated for the second entity.

[0204] Once one or both of the entities of a nominated pair have reached a threshold number of nominations, that nominated pair may be removed from a table or similar database of eligible pairs.

[0205] The automated nomination process may then proceed with the next-best ranked pair according to GRTB, and discussion topic recommendations generated for that pair. The nomination and discussion topic recommendation generation may continue until no pairs remain.

[0206] In an embodiment, HRBPs may be incorporated into a matching process by displaying a name or username of the HRBP in a user interface (such as user interface 240 described with respect to FIG. 2) with an indication that they are included in the decision to approve or reject a potential mentor mentee pair. This may provide an indication of who is involved with and / or responsible for approving a proposed match between a mentor and mentee. Knowledge of who is responsible for approving or rejecting a match may be important for determining why a potential match was rejected; including information indicating who is involved with the matching process may provide additional clarity about a strength or thoroughness of a potential mentor mentee pair.

[0207] In an embodiment, a notification may be transmitted to an appropriate entity (e.g., a mentor, a mentee, an administrator, a system, a computer, an Al agent such as a chat bot, or any suitable entity) via email, text message, phone call (such as an automated phone call), app message (e.g., a message in Slack™, Microsoft Teams™, or the like), or any suitable notification method. Notifications may be automatically generated if a mentor and / or mentee submits an application, if an administrator needs to approve a mentor or mentee application, if pairs are eligible to be nominated by a nominator, if a nominator submits match nominations, if a nominated match requires approval from an approver, if a reviewer is required to review a potential match, if a reviewer recommends that a match be rejected, if a potential match is rejected by any one or more of an approver, a mentee, or an administrator, if a potential match is approved by an approver, a mentee, or an administrator, if a proposed match needs to be reviewed and / or decided upon by a mentee, and / or when a final match is accepted by a mentee.

[0208] Notifications may provide several benefits to system users including mentors, mentees, administrators, reviewers, nominators, etc. When a pair match is proposed, but before the pair match is approved, each of the mentor and mentee may be locked until the proposed pair match is evaluated by one or more administrators, reviewers, HRBPs, and the mentee themselves. This may prevent other potential pairs having either the mentee or the mentor from being considered until the pair match is approved or rejected, which may lead to delays in providing finalized mentor-mentee pairs. The notifications provided may allow these system users to take action more quickly and reduce delays to finalizing mentor-mentee pairs.

[0209] Once a pair has been nominated to be matched, the mentor or mentee from that matched pair may be removed from further consideration if their respective quota has been reached. For example, in FIG. 1H, mentee B and mentor 4 may both have a quota of one. Once pair B / 4 is selected, all other pairs that include mentee B and mentor 4 may be temporarily locked, since the quotas for both mentee B and mentor 4 have been reached (illustrated by a lock icon next to respective pairs in FIG. 1H).

[0210] Then, the next highest, unlocked, pair may be selected. In FIG. II, the second highest ranked pair is nominated as a match, which is mentee A and mentor 1. Once again, if either (or both) of the mentor and mentee of the nominated pair reach their quota, any remaining pairs that include either the mentor or mentee from the second selected pair may be locked. FIG. II illustrates pair A / l selected to be nominated as the second pair. For example, menteeA may have a quota of one and mentor 1 may have a quota of two. Any remaining pairs including mentee A may be locked since mentee A has reached its quota, while mentor 1 remains available for nomination. For example, pair C / 1 remains eligible for nomination since mentee C has not been nominated yet and mentor 1 has a quota higher than one.

[0211] This process of nominating a mentor and mentee to be matched and locking pairs containing either the mentor or mentee where the quota has been reached for that respective mentor or mentee from the previously nominated pair may be continued until each mentee is nominated to be paired with a mentor or no unlocked eligible pairs remain. FIG. 1 J illustrates the final nomination step in which pair C / 2 is nominated to be matched. First entities and second entities may generally be matched using an analogous process, whereby a global ranking metric is created and pairs of first entities and second entities are nominated by selecting the highest ranking entity pair and locking any other pairs containing one or both entities in the nominated pair if either has reached their quota.

[0212] First entities and second entities may generally be matched using an analogous process, whereby a global ranking metric is created and pairs of first entities and second entities are matched by selecting the highest ranking entity pair and eliminating any other pairs containing one of the entities in the matched pair.

[0213] In an example, the final pairings 160 may include mentee B matched with mentor 4, mentee A matched with mentor 1, and mentee C matched with mentor 2. Once a pair has been matched, the mentor and mentee of that pair may be notified of the match. For example, a mentee match notification may be automatically sent to the mentee indicating that they should respond to the match and approve or reject the match in a timely manner. The mentee match notification may include a time limit (e.g., 3 business days, 5 business days, 10 business days, a week, etc.) until which the mentee can approve the match. If the mentee does not respond within the time limit, the match may be automatically forfeited (e.g., a match may be deleted) and different matches for the mentee and mentor proposed. In an embodiment, if a mentee fails to respond to a predetermined number of matches (e.g., 3 proposed matches in a row), the mentee may be temporarily removed from consideration for matching. A time limit for a mentee to respond and / or a predetermined number of matches to which a mentee fails to respond before being removed from consideration may be specified by an administrator, system designer, or other suitable entity.

[0214] In an aspect, a plurality of administrators may simultaneously interface with a system for mentor mentee matching in accordance with the present technology and receive updated information about changes, alterations, information added to, and information removed from an applicant profile. In some instances, the plurality of administrators may interact with (e.g., make notes about, approve, reject, etc.) the same mentor application or mentee application within a threshold amount of time. For example, if a first administrator makes a note about a mentee applicant’s profile (e.g., that the mentee should not be matched with a mentor from the legal department), the second administrator will benefit from being able to see this note automatically and in real time (and without manually reloading a user interface such as a web browser) so they do not accidentally approve a match with a mentor in the legal department. This may improve collaboration and prevent the accidental undoing of work of fellow co-administrators.

[0215] This synchronization may provide a benefit of enhancing collaboration between entities such as administrators and ensure mentor-mentee pair status is updated in real time. A processor may retrieve and store data in a plurality of tables containing mentor-mentee pair status, and the processor may further synchronize data across each of the plurality of tables. Each of the plurality of tables may include a timestamp indicating a most recent update to each of the plurality of tables. The processor may check each table at a predetermined interval; the processor may compare the timestamp indicating the most recent update of a first table to a time at which the update was last distributed across the other tables. Additionally or alternatively, the processor may periodically compare timestamps indicating most recent updates for each table of the plurality of tables. If a first table has a recent update timestamp that is more than a threshold later than the rest of the tables of the plurality of tables, the processor may determine that the first table has received new mentor-mentee pair information that has not been distributed to the other tables.

[0216] In response, the processor may update the other tables with the new mentor-mentee pair information that was sent to the first table. The updated information may then be displayed to each person or entity responsible for assessing pair compatibility and matches such as administrators.

[0217] Method 100 may include electronically scheduling a meeting 170 between the matched mentee of the plurality of mentees and the matched mentor of the plurality of eligible and compatible mentors, for example by transmitting a message related to the meeting. For example, a matched mentor-mentee pair may be notified of the match by email,text message, phone call (such as an automated phone call), app message (e.g., a message in Slack™, Microsoft Teams™, or the like), or any suitable notification method. The elimination of pairs containing previously matched entities may have the additional advantage of reducing data storage requirements. Once a pair of entities have been matched, any other potential pairs containing either of the pair of entities may be removed from memory, which has the benefit of collapsing the data storage requirements to only the data for matched pairs of entities instead of all possible matched pairs.

[0218] Electronically scheduling a meeting between the matched mentee and the matched mentor may include scheduling a meeting between matched mentee and the matched mentor (for example, electronically scheduling a meeting by sending a meeting invite via email, text message, making a calendar reservation, utilizing scheduling software, etc.), reserving a meeting location (such as a conference room, a meeting room, a desk, a restaurant, or the like), and / or providing a communications connection (such as a video call, a phone call, a link to a teleconferencing meeting, or the like).

[0219] In some instances, mentors and / or mentees may want to receive guidance on how to begin interacting with an unspecified future match or currently assigned match. Accordingly, the present technology may further include utilizing generative Al models such as LLMs, GPTs, GANs, neural networks, diffusion models, etc., to recommend conversation topics based on mentee and mentor priorities, matching criteria, skill goals, activity goals, activity distances, skill similarities, ranking metrics, mentee information, mentor information, prompts from mentors or mentees, and the like. For example, a mentee may input a natural language prompt such as the sentence “What might be a good topic to discuss with my new mentor?” to which a suitable generative Al model may provide an output of “You both have indicated an interest in developing new therapeutics for Alzheimer's disease. This may be a productive topic to discuss.” The present technology may further utilize RAG to reduce or eliminate Al model hallucinations and provide data-contextualized and tailored recommendations of discussion topics based on mentor and mentee data as outlined above.

[0220] Additionally or alternatively, a mentee and / or mentor may be provided with guidance for discussion topics automatically upon receiving final approval of the match. For example, a mentee and mentor may match based on shared interests in scientific writing, public speaking, and negotiation. Accordingly, method 100 may include providing a mentor with the following example conversation guidance, which may be generated using a suitable Al model such as an LLM: “Based on your experience in biochemistry and Mentee A’sinterest in scientific rigor, discussing scientific writing could help Mentee A improve their writing skills and learn how to clearly and concisely communicate their research findings in scientific journals, which could help Mentee A become a more effective communicator and improve their career progression.” and “Communication is a priority for Mentee A. Since public speaking is an important aspect of communication, discussing tips and techniques for public speaking - such as creating a strong narrative and engaging with the audience - could help Mentee A develop their public speaking skills.” as well as “Since Mentee A is interested in influencing / persuading others in situations where Mentee A may lack authority, discussing negotiating techniques - such as identifying interests and developing win-win scenarios - could help Mentee A become a more effective negotiator. This could help them in situations where they need to influence others, but do not have formal authority.” This guidance may be generated using contextual data (such as application data entered by mentors and / or mentees when creating an application) as well as data used to create the match itself (e.g., skill similarity, activity distance, etc.).

[0221] Method 100 may include restricting an access of a physical location based on the matching a mentee of the plurality of mentees with a mentor of the plurality of eligible and compatible mentors based on the global ranking metric. In an embodiment, restricting an access of a physical location may include controlling, activating, or deactivating a lock on a door or other barrier based on the matching. Restricting an access of a physical location may include granting access to one or more people including the matched mentor and the matched mentee based on an ID card, a token, a biometric parameter (such as a fingerprint, a retinal scan, a facial recognition, a palm print, a voice authentication, a body scan, or the like), a quick-response (QR) code, a barcode, a Rivest-Shamir-Adleman (RSA) key fob, or any suitable access element.

[0222] The method blocks as recited herein may additionally apply to matching entities beyond mentors and mentees. For example, a first plurality of entities from a first group and a second plurality of entities from a second group may be compared and matched. Each entity of the first plurality of entities may include an associated n-element first parameter vector and an associated m-element second parameter vector. Each entity of the second plurality of entities may additionally include an associated n-element first parameter vector and an associated m-element second parameter vector.

[0223] Each entity of the first plurality of entities may be compared to a plurality of eligible entities of the second plurality of entities by calculating a similarity score determined usingany suitable comparison metric of an n-element first parameter vector and an n-element first parameter vector. Comparing parameter vectors may include calculating a dot product, a Euclidean distance, an addition of two vectors, a subtraction of two vectors, a cross product, a magnitude, an average, a median, a mode, a distribution, a standard deviation, a variance, a proportion, a correlation, a regression coefficient, an analysis of variance (ANOVA) calculation, or any suitable calculation.

[0224] Each entity of the first plurality of entities may be compared to a plurality of eligible entities of the second plurality of entities by calculating a distance score determined using any suitable comparison metric of a first entity m-element second parameter vector and a second entity m-element second parameter vector, including a distance metric indicating a measure by which the second parameter vectors differ. As outlined above, comparing parameter vectors may include calculating a dot product, a Euclidean distance, an addition of two vectors, a subtraction of two vectors, a cross product, a magnitude, an average, a median, a mode, a distribution, a standard deviation, a variance, a proportion, a correlation, a regression coefficient, an analysis of variance (ANOVA) calculation, or any suitable calculation.

[0225] Each potential match may be compared from a perspective of each entity and a similarity score and distance score may be averaged for each potential pair from the perspective of each entity of the potential pair. Similarity scores and distance scores may be compared using any suitable comparison metric including any comparison metric recited above. For example, an overall average may be determined and one or more first entities from the first group may be matched with one or more second entities from the second group based on the overall average.

[0226] FIG. 2 illustrates an exemplary system 200 for performing mentor-mentee matching in accordance with the disclosure. System 200 may include a processor 210 configured to perform any or all of the method blocks recited herein. Processor 210 may include a central processing unit (CPU), a graphics processing unit (GPU), a microprocessor, a field- programmable gate array (FPGA), programmable logic controller (PLC), or any suitable circuitry configuration to perform logical operations. The processor may be communicatively coupled to memory 220, storage unit 230, and / or user interface 240.

[0227] System 200 may further include memory 220 communicatively coupled to the processor. Memory 220 may include random access memory (RAM), dynamic randomaccess memory (DRAM), flash memory, read-only memory (ROM), virtual memory, or any memory suitable for storing data and / or computer program instructions. In an embodiment, memory 220 may include software instructions to configure the processor to perform any or all of the method blocks recited herein.

[0228] Mentor information 242 and / or mentee information 244 may be stored on storage unit 230. Storage unit 230 may include a server, a hard drive (HD), a solid state drive (SSD), cloud storage, distributed storage, or any suitable information storage. Information stored on the storage unit may be updated, accessed, altered, removed, changed, relocated, or added by user interface 240, processor 210, or any suitable element.

[0229] System 200 may include at least one user interface 240 through which one or more mentors 252 and one or more mentees 254 may submit information, including mentor information 242 and / or mentee information 244. Mentor information 242 submitted by one or more mentors and mentee information 244 submitted by one or more mentees may include one or more skill goals, one or more activity goals, personal identifiable information (PII), one or more notes, one or more information related to a position within an organization (e.g., title, salary, hierarchy, department, years of experience, etc.), personal preferences, availability, and the like. A user interface 240 may be a computer, a tablet, a phone, a website, a web portal, a software application, or any suitable user interface.

[0230] FIGS. 3 A and 3B illustrate a method 300 in accordance with the present technology. Method 300 may include blocks 305-395 and may be implemented in any suitable manner, including by system 200 as instructions stored on memory 220 configuring processor 210 to perform any or all of the blocks of method 300.

[0231] Block 305 of method 300 may include generating a mentee perspective pair array comprising mentees and mentors. For example, the mentee perspective pair array may include the plurality of mentees and a plurality of eligible and compatible mentors of the plurality of mentors represented as pairs.

[0232] Block 310 of method 300 may include identifying a selected mentee. For example, a selected mentee may be identified using one or more first eligibility and compatibility criteria including years of experience, name, username, reporting structure hierarchy, department, parent organizationjob family group, job family, cost center, employee type, reporting line, mentee- or administrator-specified inclusion or exclusion criteria, threshold values or rangesof values for eligibility and compatibility parameters, or any suitable identification and / or selection criteria.

[0233] Block 315 of method 300 may include calculating a skill similarity for each potential mentor-mentee pair comprising the selected mentee and an eligible mentor. For example, a skill similarity may be calculated as a dot product of an n-element mentee skill goal vector for the selected mentee and an n-element mentor skill goal vector for the mentor of that pair to generate a skill similarity for that pair. In an embodiment, a skill similarity may be calculated as any comparison of two vectors indicating one or more skill parameters.Comparing parameter vectors may include calculating a dot product, a Euclidean distance, an addition of two vectors, a subtraction of two vectors, a cross product, a magnitude, an average, a median, a mode, a distribution, a standard deviation, a variance, a proportion, a correlation, a regression coefficient, an analysis of variance (ANOVA) calculation, or any suitable calculation.

[0234] Block 320 of method 300 may include calculating an activity distance for each pair comprising selected mentee and mentors. For example, calculating an activity distance may include calculating a Euclidean distance of an m-element mentee activity goal vector for the selected mentee and an m-element mentor activity goal vector for the mentor of that pair to generate an activity distance for that pair. Comparing activity goal vectors may include calculating a dot product, a Euclidean distance, a scalar representation, an addition of two vectors, a subtraction of two vectors, a cross product, a magnitude, an average, a median, a mode, a distribution, a standard deviation, a variance, a proportion, a correlation, a regression coefficient, an analysis of variance (ANOVA) calculation, or any suitable calculation.

[0235] Block 325 of method 300 may include generating a ranking pair comprising skill similarity and activity distance for each pair. For example, a ranking pair may include a ranking of the skill similarity and the activity distance for that pair of the mentee perspective pair array.

[0236] Block 330 of method 300 may include placing the ranking pair into a ranking group and calculating a merged mentee perspective rank.

[0237] Block 335 of method 300 may include repeating blocks 305-330 for each additional mentee of the plurality of mentees.

[0238] Block 340 of method 300 may include generating a mentor perspective pair array comprising mentees and mentors. For example, the mentor perspective pair array may includethe plurality of mentees and a plurality of eligible and compatible mentors of the plurality of mentors represented as pairs.

[0239] Block 345 of method 300 may include identifying a selected mentor. For example, a selected mentor may be identified using one or more first eligibility criteria including years of experience, name, username, reporting structure hierarchy, department, parent organization, job family group, job family, cost center, employee type, reporting line, mentor- or administrator-specified inclusion or exclusion criteria, or any suitable identification and / or selection criteria.

[0240] Block 350 of method 300 may include calculating a skill similarity for each potential mentor-mentee pair comprising the selected mentor and an eligible mentee. For example, a skill similarity may be calculated as a dot product of an n-element mentor skill goal vector for the selected mentor and an n-element mentee skill goal vector for the mentee of that pair to generate a skill similarity for that pair. In an embodiment, a skill similarity may be calculated as any comparison of two vectors indicating one or more skill parameters. Comparing parameter vectors may include calculating a dot product, a Euclidean distance, an addition of two vectors, a subtraction of two vectors, a cross product, a magnitude, an average, a median, a mode, a distribution, a standard deviation, a variance, a proportion, a correlation, a regression coefficient, an analysis of variance (ANOVA) calculation, or any suitable calculation. In an embodiment, block 350 may be skipped and the output of block 315 for the same mentee-mentor pair may be used instead.

[0241] Block 355 of method 300 may include calculating an activity distance for each pair comprising selected mentor and mentees. For example, calculating an activity distance may include calculating a Euclidean distance of an m-element mentee activity goal vector for the selected mentor and an m-element mentor activity goal vector for the mentee of that pair to generate an activity distance for that pair. Comparing activity goal vectors may include calculating a dot product, a Euclidean distance, a scalar representation, an addition of two vectors, a subtraction of two vectors, a cross product, a magnitude, an average, a median, a mode, a distribution, a standard deviation, a variance, a proportion, a correlation, a regression coefficient, an analysis of variance (ANOVA) calculation, or any suitable calculation. In an embodiment, block 355 may be skipped and the output of block 320 for the same mentee- mentor pair may be used instead.

[0242] Block 360 of method 300 may include generating a ranking pair comprising skill similarity and activity distance for each pair. For example, a ranking pair may include a ranking of the skill similarity and the activity distance for that pair of the mentor perspective pair array.

[0243] Block 365 of method 300 may include placing the ranking pair into a ranking group and calculating a merged mentor perspective rank.

[0244] Block 370 of method 300 may include repeating blocks 340-365 for each additional mentor of the plurality of mentors.

[0245] Block 375 of method 300 may include combining perspective ranks for each possible mentor-mentee pair. For example, block 375 may include combining, for each possible mentor-mentee pair from the plurality of mentees and the plurality of eligible and compatible mentors, the mentor perspective rank and the mentee perspective rank. A combining of perspective ranks may include applying a mathematical operation to the elements in each ranking pair. For example, an average, median, mode, variance, standard deviation, proportion, percentage, addition, subtraction, division, or multiplication of the elements in each ranking pair may be calculated to determine a combination.

[0246] Block 380 of method 300 may include ordering combined perspective ranks to generate a global ranking metric. For example, the combined perspective ranks may be ordered from lowest to highest, with lowest indicating a most desirable match and highest indicating a least desirable match.

[0247] Block 385 of method 300 may include nominating a mentee to be matched with a mentor based on the global ranking metric. For example, the highest-ranking mentor-mentee pair may be selected as a match.

[0248] Block 390 of method 300 may include temporarily locking pairs containing prior matched mentees and / or mentors if a pair quota for the respective mentees and / or mentors is reached. Block 390 may additionally include relinquishing a portion of memory 220 in system 200 used to store information describing pairs that are removed from consideration.

[0249] Block 395 of method 300 may include repeating blocks 385 and 390 until no additional pairs can be matched.

[0250] FIG. 4 illustrates a method 400 in accordance with the present technology. Method 405 includes blocks 405 to 425.

[0251] Block 405 may include calculating, for each potential pair comprising a selected first entity of the plurality of first entities and a selected second entity of the plurality of second entities, a similarity metric and a distance metric. A similarity metric and a distance metric may be calculated as a function of any one or more of a dot product, a Euclidean distance, a scalar representation, an addition of two vectors, a subtraction of two vectors, a cross product, a magnitude, an average, a median, a mode, a distribution, a standard deviation, a variance, a proportion, a correlation, a regression coefficient, an analysis of variance (ANOVA) calculation.

[0252] Block 410 may include ranking each potential pair relative to all other potential pairs based on the similarity metric and the distance metric for that potential pair to generate a list of ranked potential pairs. Ranking each potential pair relative to all other potential pairs based on the similarity metric and the distance metric of that potential pair to generate a list of ranked potential pairs may include applying a mathematical operation to the elements in each ranking pair. For example, an average, median, mode, variance, standard deviation, proportion, percentage, addition, subtraction, division, or multiplication of the elements in each ranking pair may be calculated to determine a ranking.

[0253] Block 415 may include selecting a top ranked pair of the list of ranked potential pairs to be matched.

[0254] Block 420 may include locking any of the other potential pairs from the list of ranked potential pairs comprising an entity with a quota that has been reached and is included in the top ranked pair to generate a list of remaining ranked potential pairs.

[0255] Block 425 may include repeating blocks 415 and 420 using the list of remaining ranked potential pairs until each of the plurality of first entities or each of the plurality of second entities has been matched.Conclusion

[0256] While various inventive embodiments have been described and illustrated herein, those of ordinary skill in the art will readily envision a variety of other means and / or structures for performing the function and / or obtaining the results and / or one or more of the advantages described herein, and each of such variations and / or modifications is deemed to be within the scope of the inventive embodiments described herein. More generally, those skilled in the art will readily appreciate that all parameters, elements, materials, and configurations described herein are meant to be exemplary and that the actual parameters,elements, materials, and / or configurations will depend upon the specific application or applications for which the inventive teachings is / are used. Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, many equivalents to the specific inventive embodiments described herein. It is, therefore, to be understood that the foregoing embodiments are presented by way of example only and that, within the scope of the appended claims and equivalents thereto, inventive embodiments may be practiced otherwise than as specifically described and claimed. Inventive embodiments of the present disclosure are directed to each individual feature, system, article, material, kit, and / or method described herein. In addition, any combination of two or more such features, systems, articles, materials, kits, and / or methods, if such features, systems, articles, materials, kits, and / or methods are not mutually inconsistent, is included within the inventive scope of the present disclosure.

[0257] Also, various inventive concepts may be embodied as one or more methods, of which an example has been provided. The acts performed as part of the method may be ordered in any suitable way. Accordingly, embodiments may be constructed in which acts are performed in an order different than illustrated, which may include performing some acts simultaneously, even though shown as sequential acts in illustrative embodiments.

[0258] All definitions, as defined and used herein, should be understood to control over dictionary definitions, definitions in documents incorporated by reference, and / or ordinary meanings of the defined terms.

[0259] The indefinite articles “a” and “an,” as used herein in the specification and in the claims, unless clearly indicated to the contrary, should be understood to mean “at least one.”

[0260] The phrase “and / or,” as used herein in the specification and in the claims, should be understood to mean “either or both” of the elements so conjoined, i.e., elements that are conjunctively present in some cases and disjunctively present in other cases. Multiple elements listed with “and / or” should be construed in the same fashion, i.e., “one or more” of the elements so conjoined. Other elements may optionally be present other than the elements specifically identified by the “and / or” clause, whether related or unrelated to those elements specifically identified. Thus, as a non-limiting example, a reference to “A and / or B”, when used in conjunction with open-ended language such as “comprising” can refer, in one embodiment, to A only (optionally including elements other than B); in another embodiment,to B only (optionally including elements other than A); in yet another embodiment, to both A and B (optionally including other elements); etc.

[0261] As used herein in the specification and in the claims, “or” should be understood to have the same meaning as “and / or” as defined above. For example, when separating items in a list, “or” or “and / or” shall be interpreted as being inclusive, i.e., the inclusion of at least one, but also including more than one, of a number or list of elements, and, optionally, additional unlisted items. Only terms clearly indicated to the contrary, such as “only one of’ or “exactly one of,” or, when used in the claims, “consisting of,” will refer to the inclusion of exactly one element of a number or list of elements. In general, the term “or” as used herein shall only be interpreted as indicating exclusive alternatives (i.e. “one or the other but not both”) when preceded by terms of exclusivity, such as “either,” “one of,” “only one of,” or “exactly one of.” “Consisting essentially of,” when used in the claims, shall have its ordinary meaning as used in the field of patent law.

[0262] As used herein in the specification and in the claims, the phrase “at least one,” in reference to a list of one or more elements, should be understood to mean at least one element selected from any one or more of the elements in the list of elements, but not necessarily including at least one of each and every element specifically listed within the list of elements and not excluding any combinations of elements in the list of elements. This definition also allows that elements may optionally be present other than the elements specifically identified within the list of elements to which the phrase “at least one” refers, whether related or unrelated to those elements specifically identified. Thus, as a non-limiting example, “at least one of A and B” (or, equivalently, “at least one of A or B,” or, equivalently “at least one of A and / or B”) can refer, in one embodiment, to at least one, optionally including more than one, A, with no B present (and optionally including elements other than B); in another embodiment, to at least one, optionally including more than one, B, with no A present (and optionally including elements other than A); in yet another embodiment, to at least one, optionally including more than one, A, and at least one, optionally including more than one, B (and optionally including other elements); etc.

[0263] In the claims, as well as in the specification above, all transitional phrases such as “comprising,” “including,” “carrying,” “having,” “containing,” “involving,” “holding,” “composed of,” and the like are to be understood to be open-ended, i.e., to mean including but not limited to. Only the transitional phrases “consisting of’ and “consisting essentiallyof’ shall be closed or semi-closed transitional phrases, respectively, as set forth in the United States Patent Office Manual of Patent Examining Procedures, Section 2111.03.

Claims

CLAIMS1. A method for creating mentor-mentee matches between a plurality of mentors and a plurality of mentees, each mentor of the plurality of mentors associated with a corresponding n-element mentor skill goal vector and a corresponding m-element mentor activity goal vector, and each mentee of the plurality of mentees associated with a corresponding n- element mentee skill goal vector and a corresponding m-element mentee activity goal vector, the method comprising:A) generating a mentee perspective pair array, the mentee perspective pair array comprising the plurality of mentees and a plurality of eligible and compatible mentors of the plurality of mentors represented as pairs;B) identifying a selected mentee of the plurality of mentees;C) calculating, for each pair of the mentee perspective pair array including the selected mentee, a dot product of an n-element mentee skill goal vector for the selected mentee and an n-element mentor skill goal vector for the mentor of that pair to generate a skill similarity for that pair;D) calculating, for each pair of the mentee perspective pair array including the selected mentee, a Euclidean distance of an m-element mentee activity goal vector for the selected mentee and an m-element mentor activity goal vector for the mentor of that pair to generate an activity distance for that pair;E) generating, for each pair of the mentee perspective pair array including the selected mentee, a ranking pair, the ranking pair comprising a ranking of the skill similarity and the activity distance for that pair of the mentee perspective pair array, and placing the ranking pair into a mentee ranking group;F) calculating, for each pair of the mentee ranking group and based on the ranking of the skill similarity and the activity distance for that pair, a merged mentee perspective rank for that pair;G) repeating B through F for each additional mentee of the plurality of mentees to generate a merged mentee perspective rank for each possible pair of mentees and mentors;H) generating a mentor perspective pair array, the mentor perspective pair array comprising the plurality of eligible and compatible mentors of the plurality of mentors and the plurality of mentees represented as pairs;I) identifying a selected mentor of the plurality of eligible and compatible mentors;J) calculating, for each pair of the mentor perspective pair array including the selected mentor, a dot product of an n-element mentor skill goal vector for the selected mentor and an n-element mentee skill goal vector for the mentee of that pair to generate a skill similarity for that pair;K) calculating, for each pair of the mentor perspective pair array including the selected mentor, a Euclidean distance of an m-element mentor activity goal vector for the selected mentor and an m-element mentee activity goal vector for the mentee of that pair to generate an activity distance for that pair;L) generating, for each pair of the mentor perspective pair array including the selected mentor, a ranking pair, the ranking pair comprising a ranking of the skill similarity and the activity distance for that pair of the mentor perspective pair array, and placing the ranking pair into a mentor ranking group;M) calculating, for each pair of the mentor ranking group and based on the ranking of the skill similarity and the activity distance for that pair, a merged mentor perspective rank for that pair;N) repeating I through M for each additional mentor of the plurality of eligible and compatible mentors to generate a merged mentor perspective rank for each possible pair of mentors and mentees;O) combining, for each possible mentor-mentee pair from the plurality of mentees and the plurality of eligible and compatible mentors, the merged mentor perspective rank and the merged mentee perspective rank to determine a global ranking metric for that mentor-mentee pair; andP) matching a mentee of the plurality of mentees with a mentor of the plurality of eligible and compatible mentors based on the global ranking metric.

2. The method of claim 1, wherein P) further comprises at least one of: transmitting a notification to an approver that a potential pair match between the mentee of the plurality of mentees and the mentor of the eligible and compatible mentors requires approval; and transmitting a notification to the mentee of the plurality of mentees indicating that they have been matched with the mentor of the plurality of eligible and compatible mentors.

3. The method of claim 2, further comprising: Q) in response to the mentee not responding to the notification within a predetermined time limit, deleting the match between the mentee and the mentor and making the mentee and the mentor available for matching again.

4. The method of claim 1, further comprising: performing a plurality of B) - F) and I) - M) in parallel, wherein performing a plurality of B) - F) and I) - M) in parallel comprises reducing a time to complete the plurality of B) - F) and I) - M) by distributing at least a portion of B) - F) and I) - M) across a plurality of processors or a plurality of cores; andP) further comprises asynchronously presenting a plurality of administrators with data associated with matching the mentee of the plurality of mentees and the mentor of the plurality of eligible and compatible mentors and receiving an approval, by at least one of the administrators, of the match between the mentee of the plurality of mentees and the mentor of the plurality of eligible and compatible mentors.

5. The method of claim 1, further comprising:R) generating, using a large language model, at least one recommendation for a discussion topic between the mentee of the plurality of mentees and the mentor of the eligible and compatible mentors based on information provided by the mentee and the mentor; andS) transmitting, to at least one of the mentee of the plurality of mentees or the mentor of the eligible and compatible mentors, the at least one recommendation for a discussion topic.

6. The method of claim 1, further comprising electronically scheduling a meeting between the matched mentee of the plurality of mentees and the matched mentor of the plurality of eligible and compatible mentors.

7. The method of claim 6, wherein electronically scheduling a meeting comprises transmitting an indication of the meeting to the matched mentee of the plurality of mentees and the matched mentor of the plurality of eligible and compatible mentors.

8. The method of claim 7, wherein electronically scheduling a meeting further comprises restricting an access of a physical location based on the matching.

9. The method of claim 8, wherein restricting the access of the physical location comprises controlling a lock.

10. The method of claim 1, wherein: the n-element mentor skill goal vectors comprise n mentor skill elements, the n mentor skill elements comprising values between 0 and 1; and the n-element mentee skill goal vectors comprise n mentee skill elements, the n mentee skill elements comprising integer values from 1 to n.

11. The method of claim 1, wherein: the m-element mentor activity goal vectors comprise m mentor activity elements, the m mentor activity elements comprising integer values from 1 to m; and the m-element mentee activity goal vectors comprise m mentee activity elements, the m mentee activity elements comprising integer values from 1 to m.

12. The method of claim 1, further comprising:Q) receiving, by a generative artificial intelligence (Al) model, first information from the mentee of the plurality of mentees or the mentor of the plurality of mentors; andR) outputting, by the generative Al model, second information in response to the first information.

13. The method of claim 12, wherein the first information comprises a prompt.

14. The method of claim 13, wherein the prompt comprises a natural language request for information associated with human conversation.

15. The method of claim 14, wherein the information associated with human conversation comprises one or more suggestions for discussion topics.

16. The method of claim 12, wherein the second information comprises one or more natural language suggestions for discussion topics.

17. The method of claim 12, wherein the generative Al model comprises at least one of a large language model (LLM), a generative pre-trained transformer (GPT), a generative adversarial network (GAN), a neural network, or a diffusion model.

18. The method of claim 12, wherein: the generative Al model utilizes retrieval augmented generation (RAG) to supplement the first information with third information, the third information received from a source other than the mentee of the plurality of mentees or the mentor of the plurality of mentors; and the generative Al model generates the second information based on the first information and the third information.

19. The method of claim 12, wherein the second information is based on at least one of personal information associated with the mentee or personal information associated with the mentor.

20. The method of claim 12, wherein the second information is based on at least one category of shared interests between the mentee and the mentor.

21. The method of claim 12, wherein the second information comprises a natural language output.

22. A non-transitory computer-readable medium containing instructions, the instructions configuring a processor to execute a method for creating mentor-mentee matches between a plurality of mentors and a plurality of mentees, each mentor of the plurality of mentors associated with a corresponding n-element mentor skill goal vector and a corresponding m- element mentor activity goal vector, and each mentee of the plurality of mentees associated with a corresponding n-element mentee skill goal vector and a corresponding m-element mentee activity goal vector, the method comprising:A) generating a mentee perspective pair array, the mentee perspective pair array comprising the plurality of mentees and a plurality of eligible and compatible mentors of the plurality of mentors represented as pairs;B) identifying a selected mentee of the plurality of mentees;C) calculating, for each pair of the mentee perspective pair array including the selected mentee, a dot product of an n-element mentee skill goal vector for the selected mentee and an n-element mentor skill goal vector for the mentor of that pair to generate a skill similarity for that pair;D) calculating, for each pair of the mentee perspective pair array including the selected mentee, a Euclidean distance of an m-element mentee activity goal vector for the selected mentee and an m-element mentor activity goal vector for the mentor of that pair to generate an activity distance for that pair;E) generating, for each pair of the mentee perspective pair array including the selected mentee, a ranking pair, the ranking pair comprising a ranking of the skill similarity and the activity distance for that pair of the mentee perspective pair array, and placing the ranking pair into a mentee ranking group;F) calculating, for each pair of the mentee ranking group and based on the ranking of the skill similarity and the activity distance for that pair, a merged mentee perspective rank for that pair;G) repeating B through F for each additional mentee of the plurality of mentees to generate a merged mentee perspective rank for each possible pair of mentees and mentors;H) generating a mentor perspective pair array, the mentor perspective pair array comprising the plurality of eligible mentors of the plurality of mentors and the plurality of mentees represented as pairs;I) identifying, based on one or more second eligibility criteria, a selected mentor of the plurality of eligible mentors;J) calculating, for each pair of the mentor perspective pair array including the selected mentor, a dot product of an n-element mentor skill goal vector for the selected mentor and an n-element mentee skill goal vector for the mentee of that pair to generate a skill similarity for that pair;K) calculating, for each pair of the mentor perspective pair array including the selected mentor, a Euclidean distance of an m-element mentor activity goal vector for the selected mentor and an m-element mentee activity goal vector for the mentee of that pair to generate an activity distance for that pair;L) generating, for each pair of the mentor perspective pair array including the selected mentor, a ranking pair, the ranking pair comprising a ranking of the skill similarityand the activity distance for that pair of the mentor perspective pair array, and placing the ranking pair into a mentor ranking group;M) calculating, for each pair of the mentor ranking group and based on the ranking of the skill similarity and the activity distance for that pair, a merged mentor perspective rank for that pair;N) repeating I through M for each additional mentor of the plurality of eligible mentors to generate a merged mentor perspective rank for each possible pair of mentors and mentees;O) combining, for each possible mentor-mentee pair from the plurality of mentees and the plurality of eligible mentors, the merged mentor perspective rank and the merged mentee perspective rank to determine a global ranking metric for that mentor-mentee pair; andP) matching a mentee of the plurality of mentees with a mentor of the plurality of eligible mentors based on the global ranking metric.

23. The medium of claim 22, wherein P) further comprises at least one of: transmitting a notification to an approver that a potential pair match between the mentee of the plurality of mentees and the mentor of the eligible and compatible mentors requires approval; and transmitting a notification to the mentee of the plurality of mentees indicating that they have been matched with the mentor of the plurality of eligible and compatible mentors.

24. The medium of claim 23, further comprising: Q) in response to the mentee not responding to the notification within a predetermined time limit, deleting the match between the mentee and the mentor and making the mentee and the mentor available for matching again.

25. The medium of claim 22, further comprising: performing a plurality of B) - F) and I) - M) in parallel, wherein performing a plurality of B) - F) and I) - M) in parallel comprises reducing a time to complete the plurality of B) - F) and I) - M) by distributing at least a portion of B) - F) and I) - M) across a plurality of processors or a plurality of cores; andP) further comprises asynchronously presenting a plurality of administrators with data associated with matching the mentee of the plurality of mentees and the mentor of theplurality of eligible and compatible mentors and receiving an approval, by at least one of the administrators, of the match between the mentee of the plurality of mentees and the mentor of the plurality of eligible and compatible mentors.

26. The medium of claim 22, further comprising:R) generating, using a large language model, at least one recommendation for a discussion topic between the mentee of the plurality of mentees and the mentor of the eligible and compatible mentors based on information provided by the mentee and the mentor; andS) transmitting, to at least one of the mentee of the plurality of mentees or the mentor of the eligible and compatible mentors, the at least one recommendation for a discussion topic.

27. The medium of claim 22, further comprising electronically scheduling a meeting between the matched mentee of the plurality of mentees and the matched mentor of the plurality of eligible and compatible mentors.

28. The medium of claim 27, wherein electronically scheduling a meeting comprises transmitting an indication of the meeting to the matched mentee of the plurality of mentees and the matched mentor of the plurality of eligible and compatible mentors.

29. The medium of claim 28, wherein electronically scheduling a meeting further comprises restricting an access of a physical location based on the matching.

30. The medium of claim 29, wherein restricting the access of the physical location comprises controlling a lock.

31. The medium of claim 22, wherein: the n-element mentor skill goal vectors comprise n mentor skill elements, the n mentor skill elements comprising values between 0 and 1; and the n-element mentee skill goal vectors comprise n mentee skill elements, the n mentee skill elements comprising integer values from 1 to n.

32. The medium of claim 22, wherein: the m-element mentor activity goal vectors comprise m mentor activity elements, the m mentor activity elements comprising integer values from 1 to m; andthe m-element mentee activity goal vectors comprise m mentee activity elements, the m mentee activity elements comprising integer values from 1 to m.

33. The medium of claim 22, further comprising:Q) receiving, by a generative artificial intelligence (Al) model, first information from the mentee of the plurality of mentees or the mentor of the plurality of mentors; andR) outputting, by the generative Al model, second information in response to the first information.

34. The medium of claim 33, wherein the first information comprises a prompt.

35. The medium of claim 34, wherein the prompt comprises a natural language request for information associated with human conversation.

36. The medium of claim 35, wherein the information associated with human conversation comprises one or more suggestions for discussion topics.

37. The medium of claim 33, wherein the second information comprises one or more natural language suggestions for discussion topics.

38. The medium of claim 33, wherein the generative Al model comprises at least one of a large language model (LLM), a generative pre-trained transformer (GPT), a generative adversarial network (GAN), a neural network, or a diffusion model.

39. The medium of claim 33, wherein: the generative Al model utilizes retrieval augmented generation (RAG) to supplement the first information with third information, the third information received from a source other than the mentee of the plurality of mentees or the mentor of the plurality of mentors; and the generative Al model generates the second information based on the first information and the third information.

40. The medium of claim 33, wherein the second information is based on at least one of personal information associated with the mentee or personal information associated with the mentor.

41. The medium of claim 33, wherein the second information is based on at least one category of shared interests between the mentee and the mentor.

42. The medium of claim 33, wherein the second information comprises a natural language output.

43. A system for creating mentor-mentee matches between a plurality of mentors and a plurality of mentees, each mentor of the plurality of mentors associated with a corresponding n-element mentor skill goal vector and a corresponding m-element mentor activity goal vector, and each mentee of the plurality of mentees associated with a corresponding n- element mentee skill goal vector and a corresponding m-element mentee activity goal vector, the system comprising: a processor; and a memory, the memory comprising instructions configuring the processor to execute a method, the method comprising:A) generating a mentee perspective pair array, the mentee perspective pair array comprising the plurality of mentees and a plurality of eligible and compatible mentors of the plurality of mentors represented as pairs;B) identifying a selected mentee of the plurality of mentees;C) calculating, for each pair of the mentee perspective pair array including the selected mentee, a dot product of an n-element mentee skill goal vector for the selected mentee and an n-element mentor skill goal vector for the mentor of that pair to generate a skill similarity for that pair;D) calculating, for each pair of the mentee perspective pair array including the selected mentee, a Euclidean distance of an m-element mentee activity goal vector for the selected mentee and an m-element mentor activity goal vector for the mentor of that pair to generate an activity distance for that pair;E) generating, for each pair of the mentee perspective pair array including the selected mentee, a ranking pair, the ranking pair comprising a ranking of the skillsimilarity and the activity distance for that pair of the mentee perspective pair array, and placing the ranking pair into a mentee ranking group;F) calculating, for each pair of the mentee ranking group and based on the ranking of the skill similarity and the activity distance for that pair, a merged mentee perspective rank for that pair;G) repeating B through F for each additional mentee of the plurality of mentees to generate a merged mentee perspective rank for each possible pair of mentees and mentors;H) generating a mentor perspective pair array, the mentor perspective pair array comprising the plurality of eligible and compatible mentors of the plurality of mentors and the plurality of mentees represented as pairs;I) identifying a selected mentor of the plurality of eligible and compatible mentors;J) calculating, for each pair of the mentor perspective pair array including the selected mentor, a dot product of an n-element mentor skill goal vector for the selected mentor and an n-element mentee skill goal vector for the mentee of that pair to generate a skill similarity for that pair;K) calculating, for each pair of the mentor perspective pair array including the selected mentor, a Euclidean distance of an m-element mentor activity goal vector for the selected mentor and an m-element mentee activity goal vector for the mentee of that pair to generate an activity distance for that pair;L) generating, for each pair of the mentor perspective pair array including the selected mentor, a ranking pair, the ranking pair comprising a ranking of the skill similarity and the activity distance for that pair of the mentor perspective pair array, and placing the ranking pair into a mentor ranking group;M) calculating, for each pair of the mentor ranking group and based on the ranking of the skill similarity and the activity distance for that pair, a merged mentor perspective rank for that pair;N) repeating I through M for each additional mentor of the plurality of eligible and compatible mentors to generate a merged mentor perspective rank for each possible pair of mentors and mentees;O) combining, for each possible mentor-mentee pair from the plurality of mentees and the plurality of eligible and compatible mentors, the merged mentor perspective rank and the merged mentee perspective rank to determine a global ranking metric for that mentor-mentee pair; andP) matching a mentee of the plurality of mentees with a mentor of the plurality of eligible and compatible mentors based on the global ranking metric.

44. The system of claim 43, wherein P) further comprises at least one of: transmitting a notification to an approver that a potential pair match between the mentee of the plurality of mentees and the mentor of the eligible and compatible mentors requires approval; and transmitting a notification to the mentee of the plurality of mentees indicating that they have been matched with the mentor of the plurality of eligible and compatible mentors.

45. The system of claim 44, further comprising: Q) in response to the mentee not responding to the notification within a predetermined time limit, deleting the match between the mentee and the mentor and making the mentee and the mentor available for matching again.

46. The system of claim 43, further comprising: performing a plurality of B) - F) and I) - M) in parallel, wherein performing a plurality of B) - F) and I) - M) in parallel comprises reducing a time to complete the plurality of B) - F) and I) - M) by distributing at least a portion of B) - F) and I) - M) across a plurality of processors or a plurality of cores; andP) further comprises asynchronously presenting a plurality of administrators with data associated with matching the mentee of the plurality of mentees and the mentor of the plurality of eligible and compatible mentors and receiving an approval, by at least one of the administrators, of the match between the mentee of the plurality of mentees and the mentor of the plurality of eligible and compatible mentors.

47. The system of claim 43, further comprising:R) generating, using a large language model, at least one recommendation for a discussion topic between the mentee of the plurality of mentees and the mentor of the eligible and compatible mentors based on information provided by the mentee and the mentor; andS) transmitting, to at least one of the mentee of the plurality of mentees or the mentor of the eligible and compatible mentors, the at least one recommendation for a discussion topic.

48. The system of claim 43, further comprising electronically scheduling a meeting between the matched mentee of the plurality of mentees and the matched mentor of the plurality of eligible and compatible mentors.

49. The system of claim 48, wherein electronically scheduling a meeting comprises transmitting an indication of the meeting to the matched mentee of the plurality of mentees and the matched mentor of the plurality of eligible and compatible mentors.

50. The system of claim 49, wherein electronically scheduling a meeting further comprises restricting an access of a physical location based on the matching.

51. The system of claim 50, wherein restricting the access of the physical location comprises controlling a lock.

52. The system of claim 43, wherein: the n-element mentor skill goal vectors comprise n mentor skill elements, the n mentor skill elements comprising values between 0 and 1; and the n-element mentee skill goal vectors comprise n mentee skill elements, the n mentee skill elements comprising integer values from 1 to n.

53. The system of claim 43, wherein: the m-element mentor activity goal vectors comprise m mentor activity elements, the m mentor activity elements comprising integer values from 1 to m; and the m-element mentee activity goal vectors comprise m mentee activity elements, the m mentee activity elements comprising integer values from 1 to m.

54. The system of claim 43, further comprising:Q) receiving, by a generative artificial intelligence (Al) model, first information from the mentee of the plurality of mentees or the mentor of the plurality of mentors; andR) outputting, by the generative Al model, second information in response to the first information.

55. The system of claim 54, wherein the first information comprises a prompt.

56. The system of claim 55, wherein the prompt comprises a natural language request for information associated with human conversation.

57. The system of claim 56, wherein the information associated with human conversation comprises one or more suggestions for discussion topics.

58. The system of claim 54, wherein the second information comprises one or more natural language suggestions for discussion topics.

59. The system of claim 54, wherein the generative Al model comprises at least one of a large language model (LLM), a generative pre-trained transformer (GPT), a generative adversarial network (GAN), a neural network, or a diffusion model.

60. The system of claim 54, wherein: the generative Al model utilizes retrieval augmented generation (RAG) to supplement the first information with third information, the third information received from a source other than the mentee of the plurality of mentees or the mentor of the plurality of mentors; and the generative Al model generates the second information based on the first information and the third information.

61. The system of claim 54, wherein the second information is based on at least one of personal information associated with the mentee or personal information associated with the mentor.

62. The system of claim 54, wherein the second information is based on at least one category of shared interests between the mentee and the mentor.

63. The system of claim 54, wherein the second information comprises a natural language output.

64. A method for matching a plurality of first entities and a plurality of second entities, the method comprising:A) calculating, for each potential pair comprising a selected first entity of the plurality of first entities and a selected second entity of the plurality of second entities, a similarity metric and a distance metric;B) ranking each potential pair relative to all other potential pairs based on the similarity metric and the distance metric for that potential pair to generate a list of ranked potential pairs;C) nominating a top ranked pair of the list of ranked potential pairs to be matched;D) locking any of the other potential pairs from the list of ranked potential pairs comprising an entity whose quota for matches has been reached to generate a list of remaining ranked potential pairs; andE) repeating C) and D) using the list of remaining ranked potential pairs until each of the plurality of first entities or each of the plurality of second entities has been matched.

65. The method of claim 64, further comprising prior to A): receiving, from at least one entity of the plurality of first entities, first information associated with the at least one entity; inputting the first information associated with the at least one entity into a generative artificial intelligence (Al) model; generating, by the generative Al model, one or more responses indicating second information associated with at least one of an improved similarity metric or an improved distance metric for at least one potential pair comprising the at least one entity; and providing the second information to the at least one entity.

66. The method of claim 65, wherein the generative Al model comprises at least one of a large language model (LLM), a generative pre-trained transformer (GPT), a generative adversarial network (GAN), a neural network, or a diffusion model.

67. The method of claim 65, wherein inputting the first information into a generative Al model comprises providing a prompt to the generative Al model.

68. The method of claim 65, wherein the second information comprises one or more suggestions for the at least one entity.

69. The method of claim 68, wherein the one or more suggestions comprise a recommendation for organizing entity-specific information used during A).

70. The method of claim 65, wherein generating, by the generative Al model, the one or more responses comprises comparing the second information to a predetermined format.

71. The method of claim 70, further comprising upon determining the one or more responses do not match the predetermined format, resubmitting the first information to the generative Al model.

72. The method of claim 65, wherein: the generative Al model utilizes retrieval augmented generation (RAG) to supplement the first information with third information, the third information received from a source other than the first entity; and the generative Al model generates the second information based on the first information and the third information.

73. The method of claim 64, wherein E) further comprises at least one of: transmitting a notification to an approver that a potential pair requires approval; and transmitting a notification to the first entity of each potential pair indicating that they have been matched with the second entity.

74. The method of claim 73, further comprising: F) in response to the first entity not responding to the notification within a predetermined time limit, forfeiting the match between the first entity and the second entity.

75. The method of claim 64, further comprising: performing A) in parallel, wherein performing A) in parallel comprises reducing a time to complete A) by distributing at least a portion of A) across a plurality of processors or a plurality of cores; andE) further comprises asynchronously presenting a plurality of administrators with data associated with matching the first entity and the second entity and receiving an approval, by at least one of the administrators, of the match between the first entity and the second entity.

76. The method of claim 64, further comprising:G) generating, using a large language model, at least one recommendation for a discussion topic between the first entity and the second entity based on information provided by the first entity and the second entity; andH) transmitting, to at least one of the first entity or the second entity, the at least one recommendation for a discussion topic.

77. The method of claim 64, wherein the similarity metric is calculated based on an average, a dot product, a variance, a standard deviation, subtraction, or addition.

78. The method of claim 64, wherein the distance metric is calculated based on a Euclidean distance, subtraction, addition, multiplication, a variance, or a standard deviation.

79. The method of claim 64, wherein the selected first entity of the plurality of first entities comprises: an n-element mentee skill goal vector comprising n mentee skill elements, the n mentee skill elements comprising integer values from 1 to n; and an m-element mentee activity goal vector comprising m mentee activity elements, the m mentee activity elements comprising integer values from 1 to m; and wherein the selected second entity of the plurality of second entities comprises: an n-element mentor skill goal vector comprising n mentor skill elements, the n mentor skill elements comprising values between 0 and 1; and an m-element mentor activity goal vector comprising m mentor activity elements, the m mentor activity elements comprising integer values from 1 to m.

80. The method of claim 64, further comprising electronically scheduling a meeting between each pair of matched entities.

81. The method of claim 80, wherein electronically scheduling a meeting comprises transmitting an indication of the meeting to each entity of each pair of matched entities.

82. The method of claim 81, wherein electronically scheduling a meeting further comprises restricting an access of a physical location based on the matching.

83. The method of claim 82, wherein restricting the access of the physical location comprises controlling a lock.

84. A non-transitory computer-readable medium containing instructions, the instructions configuring a processor to execute a method for matching a plurality of first entities and a plurality of second entities, the method comprising:A) calculating, for each potential pair comprising a selected first entity of the plurality of first entities and a selected second entity of the plurality of second entities, a similarity metric and a distance metric;B) ranking each potential pair relative to all other potential pairs based on the similarity metric and the distance metric for that potential pair to generate a list of ranked potential pairs;C) nominating a top ranked pair of the list of ranked potential pairs to be matched;D) locking any of the other potential pairs from the list of ranked potential pairs comprising an entity whose quota for matches has been reached to generate a list of remaining ranked potential pairs; andE) repeating C) and D) using the list of remaining ranked potential pairs until each of the plurality of first entities or each of the plurality of second entities has been matched.

85. The medium of claim 84, further comprising prior to A): receiving, from at least one entity of the plurality of first entities, first information associated with the at least one entity; inputting the first information associated with the at least one entity into a generative artificial intelligence (Al) model; generating, by the generative Al model, one or more responses indicating second information associated with at least one of an improved similarity metric or an improved distance metric for at least one potential pair comprising the at least one entity; and providing the second information to the at least one entity.

86. The medium of claim 85, wherein the generative Al model comprises at least one of a large language model (LLM), a generative pre-trained transformer (GPT), a generative adversarial network (GAN), a neural network, or a diffusion model.

87. The medium of claim 85, wherein inputting the first information into a generative Al model comprises providing a prompt to the generative Al model.

88. The medium of claim 85, wherein the second information comprises one or more suggestions for the at least one entity.

89. The medium of claim 88, wherein the one or more suggestions comprise a recommendation for organizing entity-specific information used during A).

90. The medium of claim 85, wherein generating, by the generative Al model, the one or more responses comprises comparing the second information to a predetermined format.

91. The medium of claim 90, further comprising upon determining the one or more responses do not match the predetermined format, resubmitting the first information to the generative Al model.

92. The medium of claim 85, wherein: the generative Al model utilizes retrieval augmented generation (RAG) to supplement the first information with third information, the third information received from a source other than the first entity; and the generative Al model generates the second information based on the first information and the third information.

93. The medium of claim 84, wherein E) further comprises at least one of: transmitting a notification to an approver that a potential pair requires approval; and transmitting a notification to the first entity of each potential pair indicating that they have been matched with the second entity.

94. The medium of claim 93, further comprising: F) in response to the first entity not responding to the notification within a predetermined time limit, deleting the match between the first entity and the second entity and making the first entity and the second entity available for matching again.

95. The medium of claim 84, further comprising:performing A) in parallel, wherein performing A) in parallel comprises reducing a time to complete A) by distributing at least a portion of A) across a plurality of processors or a plurality of cores; andE) further comprises asynchronously presenting a plurality of administrators with data associated with matching the first entity and the second entity and receiving an approval, by at least one of the administrators, of the match between the first entity and the second entity.

96. The medium of claim 84, further comprising:G) generating, using a large language model, at least one recommendation for a discussion topic between the first entity and the second entity based on information provided by the first entity and the second entity; andH) transmitting, to at least one of the first entity or the second entity, the at least one recommendation for a discussion topic.

97. The medium of claim 84, wherein the similarity metric is calculated based on an average, a dot product, a variance, a standard deviation, subtraction, or addition.

98. The medium of claim 84, wherein the distance metric is calculated based on a Euclidean distance, subtraction, addition, multiplication, a variance, or a standard deviation.

99. The medium of claim 84, wherein the selected first entity of the plurality of first entities comprises: an n-element mentee skill goal vector comprising n mentee skill elements, the n mentee skill elements comprising integer values from 1 to n; and an m-element mentee activity goal vector comprising m mentee activity elements, the m mentee activity elements comprising integer values from 1 to m; and wherein the selected second entity of the plurality of second entities comprises: an n-element mentor skill goal vector comprising n mentor skill elements, the n mentor skill elements comprising values between 0 and 1; and an m-element mentor activity goal vector comprising m mentor activity elements, the m mentor activity elements comprising integer values from 1 to m.

100. The medium of claim 84, further comprising electronically scheduling a meeting between each pair of matched entities.

101. The medium of claim 100, wherein electronically scheduling a meeting comprises transmitting an indication of the meeting to each entity of each pair of matched entities.

102. The medium of claim 101, wherein electronically scheduling a meeting further comprises restricting an access of a physical location based on the matching.

103. The medium of claim 102, wherein restricting the access of the physical location comprises controlling a lock.

104. A system for matching a plurality of first entities and a plurality of second entities, the system comprising: a processor; and a memory, the memory containing instructions configuring the processor to execute a method, the method comprising:A) calculating, for each potential pair comprising a selected first entity of the plurality of first entities and a selected second entity of the plurality of second entities, a similarity metric and a distance metric;B) ranking each potential pair relative to all other potential pairs based on the similarity metric and the distance metric for that potential pair to generate a list of ranked potential pairs;C) nominating a top ranked pair of the list of ranked potential pairs to be matched;D) locking any of the other potential pairs from the list of ranked potential pairs comprising an entity whose quota for matches has been reached to generate a list of remaining ranked potential pairs; andE) repeating C) and D) using the list of remaining ranked potential pairs until each of the plurality of first entities or each of the plurality of second entities has been matched.

105. The system of claim 104, further comprising prior to A): receiving, from at least one entity of the plurality of first entities, first information associated with the at least one entity;inputting the first information associated with the at least one entity into a generative artificial intelligence (Al) model; generating, by the generative Al model, one or more responses indicating second information associated with at least one of an improved similarity metric or an improved distance metric for at least one potential pair comprising the at least one entity; and providing the second information to the at least one entity.

106. The system of claim 105, wherein the generative Al model comprises at least one of a large language model (LLM), a generative pre-trained transformer (GPT), a generative adversarial network (GAN), a neural network, or a diffusion model.

107. The system of claim 105, wherein inputting the first information into a generative Al model comprises providing a prompt to the generative Al model.

108. The system of claim 105, wherein the second information comprises one or more suggestions for the at least one entity.

109. The system of claim 108, wherein the one or more suggestions comprise a recommendation for organizing entity-specific information used during A).

110. The system of claim 105, wherein generating, by the generative Al model, the one or more responses comprises comparing the second information to a predetermined format.

111. The system of claim 110, further comprising upon determining the one or more responses do not match the predetermined format, resubmitting the first information to the generative Al model.

112. The system of claim 105, wherein: the generative Al model utilizes retrieval augmented generation (RAG) to supplement the first information with third information, the third information received from a source other than the first entity; and the generative Al model generates the second information based on the first information and the third information.

113. The system of claim 104, wherein E) further comprises at least one of: transmitting a notification to an approver that a potential pair requires approval; and transmitting a notification to the first entity of each potential pair indicating that they have been matched with the second entity.

114. The system of claim 113, further comprising: F) in response to the first entity not responding to the notification within a predetermined time limit, forfeiting the match between the first entity and the second entity.

115. The system of claim 104, further comprising: performing A) in parallel, wherein performing A) in parallel comprises reducing a time to complete A) by distributing at least a portion of A) across a plurality of processors or a plurality of cores; andE) further comprises asynchronously presenting a plurality of administrators with data associated with matching the first entity and the second entity and receiving an approval, by at least one of the administrators, of the match between the first entity and the second entity.

116. The system of claim 104, further comprising:G) generating, using a large language model, at least one recommendation for a discussion topic between the first entity and the second entity based on information provided by the first entity and the second entity; andH) transmitting, to at least one of the first entity or the second entity, the at least one recommendation for a discussion topic.

117. The system of claim 104, wherein the similarity metric is calculated based on an average, a dot product, a variance, a standard deviation, subtraction, or addition.

118. The system of claim 104, wherein the distance metric is calculated based on a Euclidean distance, subtraction, addition, multiplication, a variance, or a standard deviation.

119. The system of claim 104, wherein the selected first entity of the plurality of first entities comprises: an n-element mentee skill goal vector comprising n mentee skill elements, the n mentee skill elements comprising integer values from 1 to n; andan m-element mentee activity goal vector comprising m mentee activity elements, the m mentee activity elements comprising integer values from 1 to m; and wherein the selected second entity of the plurality of second entities comprises: an n-element mentor skill goal vector comprising n mentor skill elements, the n mentor skill elements comprising values between 0 and 1; and an m-element mentor activity goal vector comprising m mentor activity elements, the m mentor activity elements comprising integer values from 1 to m.

120. The system of claim 104, further comprising electronically scheduling a meeting between each pair of matched entities.

121. The system of claim 120, wherein electronically scheduling a meeting comprises transmitting an indication of the meeting to each entity of each pair of matched entities.

122. The system of claim 121, wherein electronically scheduling a meeting further comprises restricting an access of a physical location based on the matching.

123. The system of claim 122, wherein restricting the access of the physical location comprises controlling a lock.

Images