Allocating electronic devices using multi-phased asynchronous data streams

Abstract

Disclosed herein are systems and associated methods for selling mobile phones and / or other used electronic devices. In some embodiments, the systems and methods can automatically allocate electronic devices (e.g., mobile phones) to a group of potential buyers. To allocate the electronic devices, the system aggregates all bids received from potential buyers and pre-negotiated prices established through purchase orders (POs). The system selects PO buyers / bidders from the group of potential buyers and allocates electronic devices to selected PO buyers / bidders by identifying opportunities for increasing revenue. To increase revenue, the system can consider, among other things, a buyer's budget constraints that were input along with their bids to determine each buyer's purchasing capacity and preferences. Additionally, if a PO buyer has committed to purchasing a certain quantity of devices at a predetermined price, the system may offer preferential treatment by prioritizing fulfilling the PO buyer's commitment before allocating additional electronic devices through the sale.

Description

ALLOCATING ELECTRONIC DEVICES USING MULTI-PHASEDASYNCHRONOUS DATA STREAMSCROSS-REFERENCE TO RELATED APPLICATION(S)

[0001] The present application claims the benefit of U.S. Provisional Patent Application No. 63 / 650,346, filed May 21 , 2024, the entirety of which is incorporated herein by reference.TECHNICAL FIELD

[0002] The present disclosure is directed generally to methods and systems for reselling electronic devices, and more particularly, to sales platforms that provide, e.g., electronic device allocation between buyers.BACKGROUND

[0003] There are more mobile phones and other electronic devices (e.g., laptop computers, notebooks, tablets, PDAs, MP3 players, wearable smart devices, etc.) in use today than there are people on the planet. The rapid growth of electronic devices is due in part to the rapid pace at which they evolve. Because of the rapid pace of development, a relatively high percentage of electronic devices are replaced every year as consumers continually upgrade to obtain the latest features or a better operating plan. As a result, many outdated or broken mobile phones and other electronic devices are simply tossed into junk drawers or otherwise kept until a suitable disposal solution arises.

[0004] Unfortunately, mobile phones and similar devices typically contain substances that can be harmful to the environment, such as arsenic, lithium, cadmium, copper, lead, mercury, and zinc. If not properly disposed of, these toxic substances can seep into groundwater from decomposing landfills and contaminate the soil with potentially harmful consequences for humans and the environment.

[0005] Some retailers take in used mobile phones via trade-in or buyback programs. As an alternative to retailer trade-in or buyback programs, consumers can recycle and / or sell their used mobile phones using self-service kiosks located in malls, retail stores, or other publicly accessible areas. Such kiosks are operated by ecoATM, LLC, the assignee of the present application, and embodiments of these kiosks are described in, forexample: U.S. Patent Nos. 8,463,946, 8,423,404, 8,239,262, 8,200,833, 8,195,811 , and 7,881 ,965, each of which is incorporated herein by reference in its entirety.BRIEF DESCRIPTION OF THE DRAWINGS

[0006] Figure 1 is a block diagram illustrating an electronic device resell process that includes manual inventory allocation.

[0007] Figure 2 is a block diagram illustrating an automated electronic device resell process configured in accordance with embodiments of the present technology.

[0008] Figures 3A-3F are a series of screenshots of one or more display pages of an online sales platform that include and / or implement aspects of the automated resell process of Figure 2, illustrating six stages of operation of the online sales platform, in accordance with embodiments of the present technology.

[0009] Figure 4 is a screenshot of one or more display pages of the online sales platform of Figures 3A-3F, illustrating features for managing buyers and users of the online sales platform, in accordance with embodiments of the present technology.

[0010] Figure 5 is a screenshot of one or more display pages of the online sales platform illustrating features for uploading inventory stored within the online sales platform, in accordance with embodiments of the present technology.

[0011] Figure 6 is a flowchart illustrating a process for allocating inventory in accordance with embodiments of the present technology.

[0012] Figure 7 is a flowchart illustrating a process for allocating inventory in accordance with embodiments of the present technology.

[0013] Figure 8 is a flowchart illustrating a process for indicating a likelihood of winning a bid in accordance with embodiments of the present technology.

[0014] Figure 9 is a block diagram illustrating an example Al system, in accordance with one or more embodiments.

[0015] Figure 10 is a block diagram illustrating an example computer system, in accordance with one or more embodiments.

[0016] The technologies described herein will become more apparent to those skilled in the art from studying the Detailed Description in conjunction with the drawings.Embodiments describing aspects of the present technology are illustrated by way of example, and the same references can indicate similar elements. While the drawings depict various embodiments for the purpose of illustration, those skilled in the art will recognize that alternative embodiments can be employed without departing from the principles of the present technologies. Accordingly, while specific embodiments are shown in the drawings, the technology is amenable to various modifications.DETAILED DESCRIPTION

[0017] Kiosks that receive electronic devices (e.g., mobile phones) for recycling typically perform various grading processes to assess a device’s condition and resale value before placing the device up for resale to potential buyers, such as wholesalers who refurbish the devices for further resale to retailers or end users. Wholesalers typically acquire bulk quantities of electronic devices from various sources such as retailers, carriers, and trade-in programs, and often specialize in specific types of devices or conditions. Wholesalers may refurbish the used electronic devices to restore their condition and / or functionality to increase the resale value and extend the lifecycle.

[0018] Some conventional approaches for reselling used electronic devices rely on labor-intensive methods such as preparing spreadsheets and manually distributing the spreadsheets to potential buyers. However, conventional methods of reselling used electronic devices can lead to delays, inconsistencies, and lost revenue. Creating and maintaining spreadsheets for inventory management, for example, can require significant manual effort and may be prone to errors, leading to inaccuracies in inventory data and potential discrepancies in pricing and product availability. Moreover, the process of sending out these spreadsheets to multiple buyers can add further complexity and consume time and resources.

[0019] An additional challenge associated with some resell processes is the lack of scalability and agility in responding to market dynamics. Manual steps can be slow and inflexible, making it difficult for the reseller to adapt quickly to changing demands, pricing trends, and inventory availability. This can result in missed opportunities to capitalize on favorable market conditions or adjust pricing strategies to increase profits effectively during the resell period (e.g., during an auction). Thus, relying on conventional methods for reselling devices can lead to lower profit margins and reduced returns on investment.

[0020] The present disclosure is generally directed to methods, apparatuses, and associated systems that provide a systematic and automated approach to selling used electronic devices. In some embodiments, an automated online platform centralizes inventory management and eliminates or at least greatly reduces the need for manual spreadsheet preparation and distribution. The automated online sales platform can allow sellers to upload inventory data and buyers to access and bid on available devices in real-time. The automated online sales platform reduces the time and effort required for managing the inventory allocation of electronic devices, enabling sellers to focus resources on other strategic aspects of their business (e.g., providing more time and resources to negotiations). By automating inventory management tasks that were previously reliant on manual spreadsheet preparation, the automated online sales platform reduces the cycle time for reselling devices and mitigates the risk of errors and inaccuracies in inventory data.

[0021] Furthermore, embodiments of the automated online sales platform described herein can provide sellers with the flexibility to adapt quickly to changing market dynamics. With automated processes and real-time data insights, sellers can respond rapidly to fluctuations in demand, adjust pricing strategies on the fly, and capitalize on emerging opportunities, ensuring that they remain competitive in a fast-paced market environment. In addition, sellers can set parameters such as quantity limits, budget constraints, and buyer preferences, and use the automated online sales platform to dynamically distribute inventory to increase profitability and return rates. Embodiments of the automated online sales platform can adapt and adjust allocation strategies in realtime based on various market dynamics, which can lead to increased realization of profitable pricing opportunities that may otherwise be missed and / or unobserved in a manually-operated platform.

[0022] In the following description, for the purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of various embodiments of the present technology. It will be apparent to one skilled in the art, however, that embodiments of the present technology can be practiced without some of these specific details.

[0023] The phrases “in some embodiments,” “in several embodiments,” “according to some embodiments,” “in the embodiments shown,” “in other embodiments,” and thelike generally mean the specific feature, structure, or characteristic following the phrase is included in at least one embodiment of the present technology and can be included in more than one embodiment. In addition, such phrases do not necessarily refer to the same embodiments or different embodiments.

[0024] Figure 1 is a block diagram illustrating an electronic device resell process 100 that includes manual inventory allocation. By way of example, the inventory can identify / include a set of smartphones, tablets, laptops, and / or other electronic and nonelectronic devices available for sale. For example, the inventory can include devices obtained through consumer-operated kiosks that meet the functional and / or cosmetic criteria for reselling (e.g., can power on, have a functioning or uncracked screen). Additionally, the inventory can include returned devices from wholesale customers. Wholesale customers may return devices from previous purchases for various reasons, such as customer returns, defects, or trade-in programs. The returned devices can be refurbished, re-graded, and added back to the inventory. Further, the inventory can include devices obtained from trade-in programs, where customers can exchange their old devices for cash or store credit. Moreover, in some instances, devices in the inventory can be directly purchased from vendors and companies. Each electronic device can be associated with device information that includes a condition of the electronic device and / or a model of the electronic device.

[0025] In block 102, the inventory may be subject to a soak period (e.g., one month) in which the inventory is held for possible reclaiming by the owner prior to resale. Additionally, the physical condition and / or electrical functionality of the electronic device may be graded (e.g., on a letter scale of A-E, etc.) during the soak period. In block 104, an automated report (e.g., an Available to Sell (ATS) report) is generated to indicate the status of the inventory. In block 106, using the ATS report, a master list of inventory ready for sale is created. The master list of inventory integrates the information from the ATS report, and serves as a reference point for electronic devices available to sell. In block 108, a barcode list for each electronic device or lot of electronic devices in the inventory (e.g., a sort list that generates an accurate list of the inventory that is ready to sell) is added to the master list.

[0026] In block 1 10, the bid sheets are created and distributed to potential buyers. In some embodiments, the bid sheets can be created by compiling information regarding the available inventory (e.g., from the master list created in block 106).

[0027] Different buyers (e.g., entities) can have different purchasing power based on a historical order / purchase volume, a geographic location, a distribution network size, and so forth. For example, buyers (e.g., large-scale wholesalers) may operate on a larger scale and source inventory from multiple channels. In some instances, these large-scale wholesalers are located outside of the United States. These buyers may prefer purchasing large quantities of devices and engage in refurbishing activities, ranging from minor repairs to complete rebuilds. Certain buyers may prefer to pay a premium for consistent availability and bulk quantities. In another example, certain medium-scale buyers may focus on triaging used devices to identify and repair those suitable for resale with minimal processing. These buyers may possess niche refurbishing capabilities and may prioritize inventory mixes with a high ratio of functional to non-functional devices. In some embodiments, these buyers value consistency in quality and volume, and may be willing to pay premium prices for inventory meeting their specific refurbishing needs. In another example, buyers, such as small wholesalers, may operate on a smaller scale than their medium or large-scale counterparts. Small wholesalers typically share similar business strategies but cater to niche distribution channels such as retail stores or online sales platforms like eBay. While small wholesalers may offer higher prices for certain inventory categories, their purchasing volumes are typically lower, reflecting their more targeted business operations.

[0028] In some embodiments, each buyer receives an individualized bid sheet tailored to the buyer’s specific preferences and requirements. The bid sheet contains information related to the inventory available for sale. For example, the bid sheet can include one row per Stock Keeping Unit (SKU) available for sale, to provide buyers with an overview of the available inventory items. Each SKU can have one or more grades (e.g., letter grades such as A, B, C, etc.) associated with it, reflecting variations in quality or condition, functionality, options, etc., which are also represented in the bid sheet. Additionally, the bid sheet can include columns where buyers can enter their bids for each SKU and grade combination. This allows buyers to specify their desired prices for the different inventory items (e.g., one or more electronic devices identified by the inventory) based on their perceived value (e.g, resources configured to be exchangedfor the inventory items) and market demand. The bid sheet can include a total estimated quantity for each SKU, aggregated across all grades. This information enables buyers to make informed decisions based on the overall availability of inventory items and helps them assess their purchasing capacity and requirements. Additionally, the bid sheet can provide bid guidance to buyers, presenting winning bids from previous sales to help buyers gauge market trends and competitive pricing and guide the buyers in formulating competitive bids aligned with current market conditions. In addition to the bid sheet, certain wholesalers (e.g., small or medium-scale wholesalers) can also receive a Quantity Limit file, which allows the wholesalers to specify the maximum quantity they wish to purchase for each SKU, grade, and / or quantity level. This allows buyers to tailor their bids according to their specific inventory needs and budget constraints, to ensure efficient allocation of resources and lower potential overspending. In some embodiments, these bid sheets are manually uploaded to a portal (e.g., by a sales operator) once the bidders submit their bids (e.g., electronic artifacts, submissions, offers, and so forth) in block 1 12. A portal refers to a user-accessible website or webpage(s) through which users can access and interact with services, tools, and content related to the electronic device resell process. A portal can include multiple webpages or sections that cater to different functionalities. For example, the webpage(s) can include dashboards, profiles, search interfaces, transactional modules, and / or informational resources.

[0029] In some embodiments, manual encumbrance checks can be performed in block 1 14, where the validity and / or status of one or more devices in the inventory are verified to ensure compliance with, e.g., regulatory and operational standards. To conduct manual encumbrance checks, the inventory is either manually reviewed and / or reviewed by automated API responses (e.g., from external Global System for Mobile Communications Association (GSMA) data providers) to queries to determine if the Individual Mobile Equipment Identifiers (IMEIs) associated with the individual devices are flagged for any encumbrances, such as having outstanding legal or financial obligations.

[0030] In block 116, following the completion of encumbrance checks, certain devices can be identified for routing to specific buyers to increase the return rate. For example, in some embodiments the inventory items that meet predetermined criteria are selected to be routed to the specific buyers, such as having clean IMEIs or satisfying specific business requirements. The criteria can vary based on factors such as device condition, model, or customer preferences. In some embodiments, once the devices areidentified, a specific list of identifiers such as barcodes representing these items is mapped to specific buyer codes. For example, all devices with clean IMEIs, verified through the encumbrance checks can be directed to a group of prioritized buyers.

[0031] When a device is identified as encumbered during the manual encumbrance check, the device is flagged and removed from the list of eligible inventory. Once the manual encumbrance checks are completed and all encumbered devices are removed, the total quantity of phones available in each lot is updated accordingly in block 118. Updating the quantity of phones available in each lot ensures that potential buyers have accurate information regarding the number of devices available for purchase. Once the quantities are updated, the seller manually assesses the pricing dynamics between purchase orders (POs) and sales prices in view of the updated quantity of phones available for sale. Additionally, an estimated winnings report that provides revenue insights into the sales outlook for the upcoming period can be manually created.

[0032] In block 124, one or more subsequent round(s) of bids can occur, allowing for potential adjustments to a buyer’s bid price and / or quantity based on, for example, market dynamics, negotiations between the reseller and buyer, and / or buyer preferences. In block 126, similar to block 112, bids are uploaded to the portal for further processing and evaluation. Subsequent bidding rounds can focus on specific buyers and product categories to increase revenue and allocate inventory more efficiently. In some embodiments, allocating inventory refers to the distribution of the electronic device inventory between different buyers. For example, allocating inventory can include determining which potential buyer receives an offer to buy particular devices in the inventory, and the quantity of particular devices a potential buyer is offered. Allocation decisions can be influenced by various factors, such as profit margins, past / forecasted demand, historical purchase / bidding patterns, market trends, customer preferences, and / or operational constraints (e.g., maintaining customer relationships). For example, to allocate inventory, the seller can collect bids from potential buyers interested in purchasing a particular device or devices in the inventory. These bids typically include information such as a quantity of the particular device or devices in the inventory that the potential buyer desires and the price offered for each item. Once the bids are received, the seller determines how best to allocate the devices in the inventory among the interested buyers. The determination can be performed manually by the seller (e.g., by sales operators of the seller).

[0033] In block 128, the process revisits the estimated winning report to refine allocation strategies and improve resource utilization. In block 130, the inventory is sorted and packed, preparing the allocated inventory for distribution to buyers in block 132. In block 134, the process certifies and adds the inventory to the designated resell channels, ensuring compliance with quality standards. The devices undergo functional testing to assess their performance and ensure that all features and functionalities are in proper working order. Any devices that do not meet the specified criteria during this testing phase are identified for further evaluation. For example, cosmetic grading can be performed to assess the physical condition of the devices, including the presence of any scratches, dents, or other signs of wear and tear.

[0034] The manual processes associated with allocating electronic devices as described above can present several challenges, particularly in segmenting buyers and products effectively. With manual segmentation, it can become cumbersome to organize separate sales for different segments of products or buyers. Manual segmentation can also lead to inefficiencies and can result in lower participation from buyers who are overwhelmed by the sheer volume of products they need to consider. Additionally, the absence of budget limits in the manual process poses a risk for buyers since, without predefined budget constraints, buyers face the challenge of managing the risk of winning more products than they can afford. Furthermore, the manual process often prioritizes bid awards solely based on the highest bidder, which can potentially lead to lower revenues due to canceled orders. Moreover, with numerous critical activities being performed manually by specific individuals, the continuity of operations may be jeopardized in the event of personnel changes or disruptions. As described below, an automated or at least partially automated sales platform configured in accordance with the present technology can reduce the dependency on manual processes, mitigating the risk of human error and / or operational disruptions.Overview of the Online Sales Platform for Automated Inventory Allocation

[0035] Figure 2 is a block diagram illustrating an automated electronic device resell process 200 configured in accordance with embodiments of the present technology. In some embodiments, portions of the process 200 are performed via one or more display pages of an online sales platform (e.g., the online sales platform 300 in Figures 3A-3F) implemented via a computer system, e.g., example computer system 1000 described inmore detail below with reference to Figure 10, in accordance with computer-executable instructions stored on a non-transitory computer-readable media. Particular features of the process 200 as embodied in certain display pages are described in more detail below with reference to Figures 3A-3F and Figure 4.

[0036] In block 202, the system includes an online sales platform (e g., the hardware and software components of the system) that automatically conducts encumbrance checks (e.g., encumbrance checks such as those discussed above with reference to Figure 1 ). For example, encumbrance checks can include IMEI-based checks for devices near the end of their soak period in block 204 against financial encumbrances. Unlike the manual encumbrance checks, automated encumbrance checks reduce the reliance on manual data entry and validation, lowering the potential for human error and ensuring greater accuracy in identifying encumbered devices. Additionally, the automated encumbrance checks provide immediate feedback on the encumbrance status of each device and allow for quicker decision-making regarding the device’s inclusion in the sales inventory. In some embodiments, the online sales platform stores the results of the encumbrance checks in the vector store discussed with reference to the inventory to facilitate easy access to historical data. One advantage of accelerating the encumbrance check process is that devices can be more quickly added to saleable inventory, increasing the opportunities for sale as compared to the manual encumbrance check process.

[0037] The information (e.g., SKUs, quantity, price) of the devices in the inventory can be stored in a cloud environment hosted by a cloud provider, or a self-hosted environment. In a cloud environment, the inventory information has the scalability of cloud services provided by platforms (e.g., AWS™, Azure™). Storing the inventory information in a cloud environment entails selecting the cloud service, provisioning resources dynamically through the provider's interface or APIs, and configuring networking components for secure communication. Cloud environments allow the inventory information to scale storage capacity without the need for manual intervention. As the demand for storage space grows, additional resources can be automatically provisioned to meet the increased workload. Additionally, cloud-based caching modules can be accessed from anywhere with an internet connection, providing convenient access to historical data for users across different locations or devices.

[0038] Conversely, in a self-hosted environment, the inventory information can be stored on a private web server. Deploying the inventory information in a self-hosted environment entails setting up the server with the necessary hardware or virtual machines, installing an operating system, and storing the inventory information. In a selfhosted environment, organizations have full control over the inventory information, allowing organizations to implement customized security measures and compliance policies tailored to the organization’s specific needs. For example, organizations in industries with strict data privacy and security regulations, such as finance institutions, can mitigate security risks by storing the inventory information in a self-hosted environment.

[0039] In block 206, the seller can define and update business rules for mapping certain electronic devices and / or groups of electronic devices to specific buyers (e.g., as described above with reference to Figure 1 ) based on, e.g., the buyers’ buyer codes. Buyer codes are used to identify buyers when they place bids, make purchases, or engage in other transactions within the sales platform. By associating each buyer with a buyer code, the platform can accurately track the buyers’ activities, monitor buyer preferences and behavior, and provide a personalized bid sheet (e.g., the bid sheet discussed in Figure 1 ). By automating the mapping, sellers can reduce manual intervention, reduce the risk of errors, and expedite the flow of inventory through the resale pipeline to the desired buyers. In some embodiments, the rules can be updated to adjust routing strategies in response to changing market conditions, inventory availability, or business requirements.

[0040] In block 208, the online sales platform automatically transfers the mapped inventory to the portal (e.g., the portal described in Figure 1 ) without the need for manual input or oversight. The automated upload ensures that approved inventory is promptly made available to the specified buyers, decreasing delays, and increasing the speed at which transactions can occur. In block 210, the inventory can be presented on a dashboard display page that allows sellers to identify particular devices in the inventory and make them available for sale / auction based on buyer preferences or market demand.

[0041] In block 212, the sales platform can automatically allocate the inventory between a group of potential buyers. Buyers are able to bid on specific quantities of inventory and set their budget for the sale. The sales platform can allow the seller tomonitor the buyer’s bidding activity and expenditure. Additionally, the sales platform can provide the sellers with allocation recommendation(s).

[0042] To generate the allocation recommendation(s), the sales platform receives input data including a purchase order (PO) price, a quantity of devices being offered, and / or a past auction sale price for similar devices. The input data is used to make decisions regarding whether to accept the offer or proceed with the sale of the devices. The received price from a bidder can be compared to the past auction sale price of the same or similar devices. The comparison can include assessing whether the difference between the received price and the past auction sale price falls within a certain margin or threshold value. For example, if the price difference is within the margin or threshold value, it can indicate to the seller that the received offer is reasonably close to the price obtained at past auctions.

[0043] In some embodiments, if the PO price is less than a threshold value but within a predefined margin, the sales platform recommends proceeding to sell the devices to the PO buyer at the offered price, since the sales platform (or, more specifically, a software algorithm included therein) can infer that the offer is competitive and aligns closely with the market value established through previous auction sales. On the other hand, if the PO is less than the threshold value by more than the preset margin, it may suggest that the offer may not be sufficiently competitive, as compared to past auction prices. In such cases, the sales platform can recommend refusing to accept the PO offer and instead proceed with auctioning the devices. By placing the devices up for auction, the sales platform can increase the potential revenue relying on the competition among potential buyers. This approach ensures that the devices are sold at prices that more closely reflect the devices’ true market value, as determined by the bidding process in the auction environment.

[0044] In some embodiments, the online sales platform selects PO buyers / bidders from the group of potential buyers and allocates inventory to the selected PO buyers / bidders by identifying opportunities for increasing revenue while achieving business objectives (e.g., maintaining business relationships). The online auction platform can consider, among other factors, the buyers’ budget constraints that were input alongside their bids to determine each buyer's purchasing capacity and preferences. For example, if a lower-priced bidder has a higher budget than a higher-priced bidder, and thus wishes to purchase a higher quantity at the lower price, the auction platform may allocate an appropriate quantify of inventor to each bidder, to maximize efficiency and profitability with respect to the sale and distribution of the devices. Additionally, the online auction platform considers pre-negotiated prices established through purchase orders (POs). For example, if a PO buyer has committed to purchasing a certain quantity of devices at a predetermined price, the platform may offer preferential treatment by prioritizing fulfilling the PO buyer’s commitment before allocating additional inventory through the auction. Despite the predetermined price being lower than that of a live bidder, the auction platform may prioritize fostering business relationships with the PO buyers. PO buyers can be invited to submit POs based on the PO buyer’s transactional history with the seller.

[0045] In block 214, the approved bids from the selected bidders on the allocated inventory are automatically recorded and uploaded to the sales platform. Upon the completion of the inventory allocation, in block 216, the inventory is sorted and packed, e.g., using methods discussed above with reference to Figure 1 , in accordance with the selected buyers to prepare the allocated inventory for distribution to buyers in block 218.

[0046] Figures 3A-3F are a series of screenshots of one or more display pages of an online sales platform 300 that include and / or implement aspects of the automated resell process of Figure 2, illustrating six stages of operation of the online sales platform, in accordance with embodiments of the present technology. The online sales platform 300 can be implemented using components a computer system, such as the example computer system 1000 described below in reference to Figure 10. In some embodiments, however, the online sales platform 300 can include different, fewer, and / or additional components or the various components can be connected in different ways.

[0047] Figure 3A illustrates an example master inventory list 302 for an upcoming sale, configured in accordance with embodiments of the present technology. The master inventory list 302 (which, e.g., may be similar to the master inventory list in block 106 of Figure 1 ), indicates available inventory items slated for sale. Each inventory item can be accompanied by device information, such as the lot identifier 304, brand 306, name 308, grade 310, and / or quantity 312, etc.

[0048] The master inventory list 302 can categorize the devices in inventory by lot, where each lot includes similar characteristics (e.g., same brand, grade, and / or name).Each lot can be assigned a unique lot identifier 304 to allow users to distinguish between different lots. The lot identifier 304 can distinguish between individual devices within the inventory list 302 by details such as brand 306, name 308, and / or grade 310. The lot identifier 304 allows buyers to identify specific inventory items based on the lot’s brand, name, and / or grade, and allows sellers to group together similar devices. Brand 306 specifies the manufacturer of each device listed in the inventory. Name 308 offers descriptive information about each device, allowing for quick identification and comparison among different inventory items. Grade 310 indicates the overall condition or grade assigned to each item based on predetermined criteria. Grade 310 helps buyers evaluate the value of individual devices and make informed decisions regarding the bidder’s bidding strategies. Moreover, quantity 312 specifies the available quantity of each inventory item (or groups of inventory items), allowing buyers to gauge the availability and potential competition for specific devices. Additionally, filter options 314 allow users to refine their search criteria and narrow down the inventory list based on specific parameters such as brand 306, name 308, and / or grade 310.

[0049] In some embodiments, the buyers are only presented with a portion of the master inventory list 302 (e.g., based on the inventory router discussed in block 206 of Figure 2). To maintain control over the participation in the auction and align with strategic objectives, the online sales platform 300 can exclude certain individuals / groups from participating in the auction altogether. For example, the online sales platform 300 can selectively restrict access to the sale based on criteria such as buyer demographics, affiliations, and / or geographic locations. For example, the online sales platform 300 can restrict the auctioning of CDMA (Code Division Multiple Access) phones to only U.S. individuals, and / or GSM (Global System for Mobiles) phones to only international buyers.

[0050] Figure 3B illustrates a display page of the of the online sales platform 300 that includes a sale scheduler 316 that facilitates the organization of sales for the master inventory 302. In some embodiments, the online sales platform 300 can enable the partitioning of sales into one or more segmented rounds 318.Each round 318 in the sale can be scheduled independently for specific times or days. The scheduling functionality allows sellers to plan and coordinate the sales events according to preferences and business needs. Further, the online sales platform 300 can incorporate rules 320 for each sale, allowing sellers to tailor the sale parameters to suitspecific requirements. For example, sellers can define rules 320 specifying that only a portion of the master inventory 302 is to be sold in a particular round.

[0051] Figure 3C illustrates a display page for submitting a bid via the online sales platform 300. In some embodiments, a bid submission 322 can be submitted by designated individuals other than the bidder, such as sales representatives, but on behalf of the bidder. In some embodiments, the online sales platform 300 allows the sales representatives to change the bid price directly through the online sales platform 300 after negotiating with the bidder.

[0052] In some implementations, the seller (e.g., a selling organization) approves of the bids submitted by the bidder and / or the sales representatives of the seller. In some implementations, the online sales platform 300 implements an artificial intelligence (Al) model (e.g., an Al assistant) for anomaly detection to aid the seller (e.g., a sales team of the seller) in approving bids. As bids are submitted, the Al model can monitor the bidding activity and evaluate various parameters such as bid amounts, bidder behavior patterns, and historical data. The Al model learns to identify anomalies or deviations from typical bidding behavior, flagging bids that may warrant closer scrutiny by the sales team. For example, the Al model can be trained on historical bidding data to detect various types of anomalies, including outliers, sudden changes in behavior, and suspicious patterns. Methods of training an Al model are described further with respect to Figure 9. By processing large volumes of bid data in real-time, the Al model can detect anomalies that might otherwise go unnoticed, such as unusually high or low bid amounts, rapid changes in bidding behavior, or bid patterns inconsistent with historical norms. Using an Al model to detect potential anomalies can help to prevent fraudulent or suspicious bidding activity.

[0053] Figure 3D illustrates a display page for confirming a bid submittal via the online sales platform 300. A bid submission confirmation 324 is displayed when bidders confirm their bid submissions. To confirm their bid submissions, bidders are prompted to input their bid amounts for specific lots they wish to acquire. The interface includes fields where bidders can enter their bid amounts alongside the corresponding lot numbers or identifiers. A download option 326 can be displayed alongside the bid confirmation 324, allowing bidders to download a copy of their bid submissions for record-keeping purposes. By providing a download option, the online sales platform allows bidders and sellers to track the bidder’s bids and reference them as needed. In some embodiments,the bid is a sealed bid phase where buyers submit their offers (including price, quantity, and / or type of device) to the online sales platform without knowledge of other buyers' bids or the current leading bid.

[0054] Figure 3E illustrates a dynamic sales report of the online sales platform 300. The dynamic sales reports can include columns 328, 330, that provide information regarding the performance of specific identifiers or lots within the sale, allowing administrators to assess profitability and strategic opportunities. For example, the columns 328, 330 can include metrics and data points relevant to each particular identifier or lot up for sale. For example, column 328 can include a revenue metric which indicates the total expected income generated from the sale of a specific lot (or portion of the lot) based on the current bidding activity and pricing dynamics. Revenue metrics allow sellers to gauge the potential financial returns associated with the sale of a particular device or lot of devices, helping them prioritize and allocate their resources strategically. In addition to revenue, the dynamic sale reports can include a margin metric in column 330 which represents the difference between the revenue generated from the sale of a lot and the associated costs or expenses. The margin metric in column 330 offers profitability insights, enabling sellers to assess the profitability of selling particular devices or lots relative to the seller’s cost structures and financial objectives. By understanding the margin potential of each identifier, sellers can make informed decisions that align with their profitability targets and business strategies. Automating the calculation of margin data can also accelerate the calculation process and provide the sellers with more opportunities to improve the margins based on adjustments to parameters within the auction process.

[0055] Figure 3F illustrates a dynamic sale report of the online sales platform 300. As the bidding process progresses and new bids are submitted, the sale reports undergo revisions and updates to reflect the latest bidding activity and pricing dynamics, providing participants with real-time insights into the evolving market conditions and competitive landscape. The dynamic sale report includes columns 328, 330 which capture the fluctuations and changes in key metrics and data points following an additional round of bidding. For example, in Figure 3F, after Round 2 is complete, the quantity of each lot as well as the average bid for each lot is updated on the sales platform 300.

[0056] Figure 4 is a screenshot of one or more display pages of an online sales platform (e.g., the online sales platform 300 in Figures 3A-3F), illustrating features for managing buyers and users of the online sales platform, in accordance with embodiments of the present technology. The screenshot can include a buyer management module 402, buyer code field 404, existing buyer field 406, and existing buyer code field 408.

[0057] The buyer management module 402 can organize the various buyers and users participating in the online sales platform. Buyer code field 404 serves as a unique identifier assigned to each buyer within the online sales platform, facilitating the identification and tracking of individual users. The existing buyer field 406 provides administrators with a detailed overview of all registered buyers, allowing them to view essential information such as user profiles, contact details, transaction history, and the existing buyer code field 408. The buyer code field 404, 408 can categorize buyers based on predefined criteria or attributes, allowing sellers to implement targeted management strategies and tailor their approach to different buyer segments effectively. By grouping buyers according to specific criteria, such as purchasing behavior or geographic location, sellers can better select bidders and allocate the inventory to increase the return rate.

[0058] Figure 5 is a screenshot of one or more display pages of the online sales platform illustrating features for uploading inventory stored within the online sales platform, in accordance with embodiments of the present technology. The screenshot includes an inventory management module 502, inventory upload interface 504, inventory selection interface 506, and particular time period 508.

[0059] The inventory management module 502 can allow a seller to manage inventory-related activities within the sales platform. Within the inventory management module 502, an inventory upload interface 504 can provide users with options to upload inventory files for a specific time period, such as a particular week. For example, the inventory upload interface 504 can be a web form with a browse and / or an upload button that allows users to select and upload inventory files from their devices. An inventory selection interface 506 can enable users to navigate between different weeks or other time periods to view inventory data. For example, the inventory selection interface 506 can include a graphical user interface (GUI) with dropdown menus or calendar controls that allow users to select a desired time frame (e.g., a particular week). In someembodiments, a search functionality can be included in the GUI to enable users to quickly locate specific inventory data based on custom search criteria. When a user selects the inventory of a particular time period 508, the inventory management module 502 can visually highlight the indicator indicating the particular time period 508. For example, the inventory management module 502 can include graphical elements such as color coding or highlighting to distinguish the particular time period 508 from other time periods.

[0060] Figure 6 is a flowchart illustrating a process 600 for allocating inventory in accordance with embodiments of the present technology. In some embodiments, the process 600 is performed via one or more display pages / interfaces of the online sales platform (e.g., the online sales platform 300 in Figures 3A-3F) implemented via a computer system, e.g., the example computer system 1000 described in more detail below with reference to Figure 10.

[0061] In block 602, the online sales platform can cause a display (e.g., via a user interface of a computing device) of an inventory identifying a set of electronic devices each associated with device information. The device information can include, for example, a unique identifier, a name, a source identifier, an associated inventory quantity (e.g., an inventory quantity associated with a particular inventory in which the device is included in), a condition, a grade, a carrier, a model, and / or other attributes of the electronic device. In some embodiments, the device information can be stored in a cloud environment hosted by a cloud provider and / or a self-hosted environment.

[0062] In block 604, the online sales platform receives (e.g., via the user interface) a user request representing a set of electronic artifacts (e.g., a bid, an electronic submission) that each includes an identification of (i) one or more electronic devices within the set of electronic devices and (ii) a set of resources (e.g., a price, monetary value, asset) configured to be exchanged for the one or more electronic devices. Each electronic artifact can be linked to an entity associated with a set of attributes that include a historical order volume, a geographic location, a distribution network size, and / or other attributes of the entity. The user request can be stored in a cloud environment hosted by a cloud provider and / or a self-hosted environment.

[0063] In block 606, the online sales platform uses (i) the set of electronic artifacts, (ii) respective device information of each electronic device in the set of electronic devices, and / or (iii) the set of attributes of the entity linked to each electronic artifact todetermine / generate a subset of electronic artifacts (e.g., the “winners” of the auction). For example, as described below in Figure 8 and 9, the online sales platform can, responsive to receiving the user request, transmit an input including the set of electronic artifacts, respective device information of each electronic device in the set of electronic devices, and / or the set of attributes of the entity linked to each electronic artifact into an Al model trained to generate an output including an identification of a subset of electronic artifacts and / or allocated electronic devices for each of the subset of electronic artifacts (e.g., allocation recommendation(s)). The Al model can be trained using historical allocation data that includes previously received electronic artifacts corresponding previously allocated electronic devices, historical sets of electronic devices, one or more patterns associated with the linked entities, and so forth. The online sales platform can obtain (e.g., receive via the user interface) a user selection approving one or more allocation recommendations from the output, and initiate one or more actions to distribute the allocated electronic devices of each of the subset of electronic artifacts to respective entities.

[0064] The Al model in Figures 6, 8, and / or 9 can be a neural network with an encoder-decoder architecture trained to encode the user request into a first set of latent space representations. The Al model (which can be an aggregated model framework of multiple models) can compare the first set of latent space representations with a second set of latent space representations derived from the set of electronic devices to determine the allocated electronic devices for each of the subset of electronic artifacts, and decode respective latent space representations of the subset of electronic artifacts to generate the output.

[0065] To determ ine / generate the subset of electronic artifacts, the online sales platform can determine, for each electronic artifact in the set of electronic artifacts, a ratio between the set of resources and a quantity of electronic devices in the one or more electronic devices. In some embodiments, the online sales platform can assign, for each electronic device in the set of electronic devices, a weight to the entity linked to each electronic artifact based on (i) the device information of each electronic device and (ii) the set of attributes of the entity. The online sales platform can allocate each electronic device in the set of electronic devices to one or more electronic artifacts in accordance with (i) respective ratios of the one or more electronic artifacts and / or (ii) respective assigned weight of the entity linked to the one or more electronic artifacts. For example,the online sales platform ranks the set of electronic artifacts based on respective ratios and assigned weights, and subsequently allocates the electronic device to a particular electronic artifact having a highest (or top k) rank. To determine the weight, the online sales platform can generate a numerical score for each attribute in the set of attributes and aggregate the numerical score for each attribute in the set of attributes.

[0066] In some embodiments, the online sales platform selects the highest bidder. For example, if the online sales platform determines that a first electronic artifact and a second electronic artifact both specify a common electronic device, the online sales platform can compare a first monetary value of the first electronic artifact to a second monetary value of the second electronic artifact (i.e., comparing respective sets of resources), and allocate the common electronic device to the electronic artifact having a higher monetary value (i.e., a larger set of resources).

[0067] The online sales platform can receive (e.g., via the user interface) a set of allocation constraints including a maximum quantity of electronic devices per entity, a minimum quantity of electronic devices per entity, and / or a device type restriction per entity, and can allocate the electronic devices in accordance with the set of allocation constraints. For example, when a first electronic artifact and a second electronic artifact have equal monetary values (i.e., equal sets of resources), the online sales platform can compare a first maximum quantity of the first electronic artifact to a second maximum quantity of the second electronic artifact, and allocate an associated electronic device to the electronic artifact having a higher maximum quantity (i.e., allocating the devices to a bidder that bids a higher quantity).

[0068] In some embodiments, one or more user requests indicate a budget or budget range (i.e., a maximum resource allocation limit) across one or more bids. The online sales platform can track a cumulative resource allocation (i.e., the resources corresponding to electronic devices allocated or bid upon) for the particular entity across the multiple bids (e.g., multiple electronic artifacts), and compare the cumulative resource allocation to the maximum resource allocation limit. Responsive to the cumulative resource allocation exceeding the maximum resource allocation limit, the online sales platform can prevent allocation of subsequently allocated electronic devices of the set of electronic devices to the particular entity.

[0069] In block 608, the online sales platform causes a display (e.g., via the user interface) of a representation of the subset of electronic artifacts and / or the allocated electronic devices for each of the subset of electronic artifacts (e.g., an acceptance status for one or more bids such as accept, decline, counteroffer, and so forth). In some implementations, the online sales platform generates a report indicating the subset of electronic artifacts. The report can include a description indicating a respective ratio and / or the assigned weight for each allocation. The report can be stored in a cloud environment hosted by a cloud provider and / or a self-hosted environment. To notify the bidders / entities, the online sales platform can transmit, to one or more entities associated with the subset of electronic artifacts, a notification indicating acceptance of the subset of electronic artifacts. In some embodiments, the online sales platform can automatically or manually (e.g., responsive to subsequent user input / approval) initiate a transaction for each electronic artifact in the subset of electronic artifacts based on the subset of electronic artifacts.

[0070] In auctions with multiple bidding rounds, the online sales platform can receive (e.g., via the user interface) a second user request representing a second set of electronic artifacts that each includes an identification of a second set of electronic devices (e.g., one or more) within the set of electronic devices and / or a second set of resources configured to be exchanged for the second set of electronic devices. In some embodiments, the bidders are only enabled to bid equal to or higher than that of the previous round (i.e., so that the second set of resources is larger than the first set of resources). The online sales platform can determine a first allocation of electronic devices based on the first set of electronic artifacts and cause a display (e.g., via the user interface) of the first allocation.

[0071] The online sales platform can determine a second allocation of electronic devices based on bids from multiple rounds (e.g., the first set of electronic artifacts, the second set of electronic artifacts, and so forth). Allocations from subsequent bidding rounds can be performed using an Al model. For example, responsive to receiving the second user input, the online sales platform can input the second set of electronic artifacts, respective device information of each electronic device in the second set of electronic devices, and / or a respective set of attributes of a respective entity linked to each of the second set of electronic artifacts into the Al model trained to output anidentification of a second subset of electronic artifacts and / or allocated electronic devices for each of the second subset of electronic artifacts.Example Methods of Using Al Models Within the Online Sales Platform

[0072] Figure 7 is a flowchart illustrating a process 700 for allocating inventory in accordance with embodiments of the present technology. In some embodiments, the process 700 is performed via one or more display pages of an online sales platform (e.g., the online sales platform 300 in Figures 3A-3F) implemented via a computer system, e.g., the example computer system 1000 described in more detail below with reference to Figure 10.

[0073] In block 702, the online sales platform provides an inventory including one or more sets of electronic devices. By way of example, the inventory can include used smartphones, tablets, laptops, or other electronic gadgets available for auction or sale. Each group (e.g., each lot) within the inventory represents a distinct grouping of electronic devices, which may be grouped based on brand, model, condition, quantity, and other attributes of the electronic devices. The online sales platform can cause a display, via a user interface of a computing device, of the inventory identifying the electronic devices. Each can be associated with device information that includes a brand, model, condition, quantity, and other attributes of the electronic devices.

[0074] In block 704, the online sales platform obtains training data including a set of historical data (e.g., bidding data, purchase order data), a set of buyer budget data, and / or a set of revenue data. Historical bidding data captures the bidding behavior of buyers during previous sales. Historical purchase order data details past purchase orders placed by buyers and can include details such as the types of inventories purchased and the quantities ordered. Budget data outlines the budgetary constraints and spending limits of individual buyers. For example, the budget data can be obtained by the budgets input by buyers when bidding, such as in block 212 of Figure 2. Revenue data outlines the revenue outcomes of previous allocation decisions.

[0075] In block 706, based on the training data, the online sales platform obtains and / or trains an Al model to dynamically allocate the sets of electronic devices of the inventory. In some embodiments, the online sales platform preprocesses the training data to clean and format the data before inputting the data into the Al model. Thepreprocessing step can include handling missing values, normalizing data, and encoding categorical variables to prepare the training dataset for analysis.

[0076] During the training process, the Al model can identify underlying patterns and relationships between various factors, such as bidder behavior, historical performance, market demand, and revenue optimization objectives. Through this training process, the Al model develops an understanding of how to allocate the inventory effectively, increasing the revenue return rate while satisfying allocation constraints and objectives (e.g., allocating devices to PO buyers if the difference in the PO price and the bid price is within a certain threshold). The Al model can adjust its parameters to lower prediction errors and increase its ability to allocate inventory while achieving an increased return rate. In some embodiments, the trained Al model can be one or more of: decision trees, random forests, gradient-boosting machines, or neural networks.

[0077] In block 708, based on the trained Al model, the online sales platform generates one or more selected buyers (or buyer codes) for each set of the electronic devices in the inventory. Based on the learned parameters, the Al model dynamically assigns buyers to specific sets of electronic devices based on the Al model’s predicted allocation with an increased return rate. The allocation decisions made by the Al model aim to increase the return rate by matching each set of electronic devices with the most suitable buyers, decreasing the likelihood of the allocated buyer canceling orders, and increasing overall profitability.

[0078] In block 710, the online sales platform presents an indicator that indicates the selected buyers for each set of electronic devices of the inventory. The indication can be a visual representation of the allocation decisions made by the Al model, to allow users to review and understand the reasoning behind each assignment. The presentation can take various forms, such as a report, GUI, or interactive dashboard. In some embodiments, sellers can have the opportunity to review and adjust the allocations based on preferences and strategic objectives before finalizing the distribution of electronic devices to buyers.

[0079] For example, the online sales platform receives an input that includes (i) the electronic artifacts / bids, (ii) the respective device information of each electronic device, and / or (iii) the set of attributes of the entity / buyer linked to each electronic artifact. The online sales platform can be trained to output an identification of a subset of bids (i.e.,select the buyers) by determining, for each electronic artifact in the set of electronic artifacts, a ratio between the set of resources and a quantity of electronic devices in the one or more electronic devices. The online sales platform assigns, for each electronic device in the set of electronic devices, a weight to the entity linked to each electronic artifact based on (i) the device information of the electronic device and (ii) the set of attributes of the entity, and allocates each electronic device in the set of electronic devices to one or more electronic artifacts in accordance with (i) respective ratios of the one or more electronic artifacts and (ii) respective assigned weight of the entity linked to the one or more electronic artifacts.

[0080] Figure 8 is a flowchart illustrating a process 800 for indicating a likelihood of winning a bid in accordance with embodiments of the present technology. In some embodiments, the process 800 is performed via one or more display pages of an online sales platform (e.g., the online sales platform 300 in Figures 3A-3F) implemented via a computer system, e.g, the example computer system 1000 described in more detail below with reference to Figure 10. Particular aspects of the online sales platform are illustrated and described in more detail above with reference to Figures 3A-3F, Figure 4, and Figure 5. Other embodiments can include different, fewer, and / or additional steps or can perform the steps in different orders.

[0081] In block 802, the online sales platform (e.g., the online sales platform of Figures 3A-3F, Figure 4, and Figure 5) can provide a historical dataset containing past sales data. The past sales data can include a quality, a quantity, a price, and / or a condition of one or more winning bidders.

[0082] In block 804, the online sales platform supplies an Al model with a price and a quantity (e.g., a bidder’s maximum quantity) of a bid, wherein the Al model is trained on the historical dataset. In block 806, the online sales platform receives, from the Al model, the likelihood of the bid being the winning bid. As buyers enter their bids, the Al model can evaluate factors such as the quantity, quality, and condition of the items subject to bidding, as well as historical bidding patterns and outcomes. As the online sales platform executes the sales, the Al model can collect data related to the evaluated factors and continuously learns from past auction data to provide a likelihood of winning a bid for a particular bid price and / or bid quantity. Methods of training an Al model are described further with respect to Figure 9. Buyers can adjust their bid amounts based onthe likelihood of winning to increase their chances of success while staying within their budget constraints. In some embodiments, after the first round of bidding, bidders are unable to decrease their bids in subsequent rounds, but can change their bid by increasing it.

[0083] In some implementations, the likelihood of winning a bid can be determined by referencing a predetermined ruleset. For example, a ruleset can include a range of bids and a range of corresponding likelihoods of winning a bid. The range of bids and the range of corresponding likelihoods of winning a bid can be manually input into the online sales platform. In block 808, the online sales platform presents, through the sales platform, an indication of the likelihood of the bid being the winning bid. The online sales platform can indicate (e.g., via a visual graphical indicator) to a bidder a likelihood of winning a bid.Embodiments of Al Systems

[0084] Figure 9 is a high-level block diagram illustrating an example Al system configured in accordance with one or more embodiments of the present technology. The Al system 900 is implemented using components of the example computer system 1000 illustrated and described in more detail with reference to Figure 10. Likewise, embodiments of the Al system 900 include different and / or additional components or be connected in different ways.

[0085] In some embodiments, as shown in Figure 9, the Al system 900 includes a set of layers, which conceptually organize elements within an example network topology for the Al system’s architecture to implement a particular Al model 930. Generally, an Al model 930 is a computer-executable program implemented by the Al system 900 that analyses data to make predictions. Information passes through each layer of the Al system 900 to generate outputs for the Al model 930. The layers include a data layer 902, a structure layer 904, a model layer 906, and an application layer 908. The algorithm 916 of the structure layer 904 and the model structure 920 and model parameters 922 of the model layer 906 together form the example Al model 930. The optimizer 926, loss function engine 924, and regularization engine 928 work to refine and optimize the Al model 930, and the data layer 902 provides resources and support for the application of the Al model 930 by the application layer 908.

[0086] The data layer 902 acts as the foundation of the Al system 900 by preparing data for the Al model 930. As shown, in some embodiments, the data layer 902 includes two sub-layers: a hardware platform 910 and one or more software libraries 912. The hardware platform 910 is designed to perform operations for the Al model 930 and includes computing resources for storage, memory, logic, and networking, such as the resources described in relation to Figures 1-8. The hardware platform 910 processes amounts of data using one or more servers. The servers can perform backend operations such as matrix calculations, parallel calculations, machine learning (ML) training, and the like. Examples of servers used by the hardware platform 910 include central processing units (CPUs) and graphics processing units (GPUs). CPUs are electronic circuitry designed to execute instructions for computer programs, such as arithmetic, logic, controlling, and input / output (I / O) operations, and can be implemented on integrated circuit (IC) microprocessors. GPUs are electric circuits that were originally designed for graphics manipulation and output but may be used for Al applications due to vast computing and memory resources. GPUs use a parallel structure that generally makes processing more efficient than that of CPUs. In some instances, the hardware platform 910 includes Infrastructure as a Service (laaS) resources, which are computing resources, (e.g., servers, memory, etc.) offered by a cloud services provider. In some embodiments, the hardware platform 910 includes computer memory for storing data about the Al model 930, application of the Al model 930, and training data for the Al model 930. In some embodiments, the computer memory is a form of random-access memory (RAM), such as dynamic RAM, static RAM, and non-volatile RAM.

[0087] In some embodiments, the software libraries 912 may be suites of data and programming code, including executables, used to control the computing resources of the hardware platform 910. In some embodiments, the programming code includes low- level primitives (e.g., fundamental language elements) that form the foundation of one or more low-level programming languages, such that servers of the hardware platform 910 can use the low-level primitives to carry out specific operations. The low-level programming languages do not require much, if any, abstraction from a computing resource’s instruction set architecture, allowing them to run quickly with a small memory footprint. Examples of software libraries 912 that can be included in the Al system 900 include Intel Math Kernel Library, Nvidia cuDNN, Eigen, and Open BLAS.

[0088] In some embodiments, the structure layer 904 includes an ML framework 914 and an algorithm 916. The ML framework 914 can be thought of as an interface, library, or tool that allows users to build and deploy the Al model 980. In some embodiments, the ML framework 914 includes an open-source library, an application programming interface (API), a gradient-boosting library, an ensemble method, and / or a deep learning toolkit that works with the layers of the Al system facilitate development of the Al model 930. For example, the ML framework 914 distributes processes for the application or training of the Al model 930 across multiple resources in the hardware platform 910. In some embodiments, the ML framework 914 also includes a set of prebuilt components that have the functionality to implement and train the Al model 930 and allow users to use pre-built functions and classes to construct and train the Al model 930. Thus, the ML framework 914 can be used to facilitate data engineering, development, hyperparameter tuning, testing, and training for the Al model 930. Examples of ML frameworks 914 that can be used in the Al system 900 include TensorFlow, PyTorch, Scikit-Learn, Keras, Caffe, LightGBM, Random Forest, and Amazon Web Services.

[0089] In some embodiments, the algorithm 916 is an organized set of computerexecutable operations used to generate output data from a set of input data and can be described using pseudocode. In some embodiments, the algorithm 916 includes complex code that allows the computing resources to learn from new input data and create new / modified outputs based on what was learned. In some embodiments, the algorithm 916 builds the Al model 930 through being trained while running computing resources of the hardware platform 910. The training allows the algorithm 916 to make predictions or decisions without being explicitly programmed to do so. Once trained, the algorithm 916 runs at the computing resources as part of the Al model 930 to make predictions or decisions, improve computing resource performance, or perform tasks. The algorithm 916 is trained using supervised learning, unsupervised learning, semi-supervised learning, and / or reinforcement learning. The application layer 908 describes how the Al system 900 is used to solve problems or perform tasks.

[0090] As an example, to train an Al model 930 that is intended to model human language (also referred to as a language model), the data layer 902 is a collection of text documents, referred to as a text corpus (or simply referred to as a corpus). The corpus represents a language domain (e.g., a single language), a subject domain (e.g., scientific papers), and / or encompasses another domain or domains, be they larger or smaller than- J-a single language or subject domain. For example, a relatively large, multilingual, and non-subject-specific corpus is created by extracting text from online web pages and / or publicly available social media posts. In some embodiments, data layer 902 is annotated with ground truth labels (e.g., each data entry in the training dataset is paired with a label), or unlabeled.

[0091] Training an Al model 930 generally involves inputting into an Al model 930 (e.g., an untrained ML model) data layer 902 to be processed by the Al model 930, processing the data layer 902 using the Al model 930, collecting the output generated by the Al model 930 (e.g., based on the inputted training data), and comparing the output to a desired set of target values. If the data layer 902 is labeled, the desired target values, in some embodiments, are, e.g., the ground truth labels of the data layer 902. If the data layer 902 is unlabeled, the desired target value is, in some embodiments, a reconstructed (or otherwise processed) version of the corresponding Al model 930 input (e.g., in the case of an autoencoder), or is a measure of some target observable effect on the environment (e.g., in the case of a reinforcement learning agent). The parameters of the Al model 930 are updated based on a difference between the generated output value and the desired target value. For example, if the value outputted by the Al model 930 is excessively high, the parameters are adjusted so as to lower the output value in future training iterations. An objective function is a way to quantitatively represent how close the output value is to the target value. An objective function represents a quantity (or one or more quantities) to be optimized (e.g., minimize a loss or maximize a reward) in order to bring the output value as close to the target value as possible. The goal of training the Al model 930 typically is to minimize a loss function or maximize a reward function.

[0092] In some embodiments, the data layer 902 is a subset of a larger data set. For example, a data set is split into three mutually exclusive subsets: a training set, a validation (or cross-validation) set, and a testing set. The three subsets of data, in some embodiments, are used sequentially during Al model 930 training. For example, the training set is first used to train one or more ML models, each Al model 930, e.g., having a particular architecture, having a particular training procedure, being describable by a set of model hyperparameters, and / or otherwise being varied from the other of the one or more ML models. The validation (or cross-validation) set, in some embodiments, is used as input data into the trained ML models to, e.g., measure the performance of the trained ML models and / or compare performance between them. In some embodiments,where hyperparameters are used, a new set of hyperparameters is determined based on the measured performance of one or more of the trained ML models, and the first act of training (i.e., with the training set) begins again on a different ML model described by the new set of determined hyperparameters. The steps are repeated to produce a more performant trained ML model. Once such a trained ML model is obtained (e.g., after the hyperparameters have been adjusted to achieve a desired level of performance), a third act of collecting the output generated by the trained ML model applied to the third subset (the testing set) begins in some embodiments. The output generated from the testing set, in some embodiments, is compared with the corresponding desired target values to give a final assessment of the trained ML model’s accuracy. Other segmentations of the larger data set and / or schemes for using the segments for training one or more ML models are possible.

[0093] Backpropagation is an algorithm for training an Al model 930. Backpropagation is used to adjust (also referred to as update) the value of the parameters in the Al model 930, with the goal of optimizing the objective function. For example, a defined loss function is calculated by forward propagation of an input to obtain an output of the Al model 930 and a comparison of the output value with the target value. Backpropagation calculates a gradient of the loss function with respect to the parameters of the ML model, and a gradient algorithm (e.g., gradient descent) is used to update (i.e., “learn”) the parameters to reduce the loss function. Backpropagation is performed iteratively so that the loss function is converged or minimized. In some embodiments, other techniques for learning the parameters of the Al model 930 are used. The process of updating (or learning) the parameters over many iterations is referred to as training. In some embodiments, training is carried out iteratively until a convergence condition is met (e.g., a predefined maximum number of iterations has been performed, or the value outputted by the Al model 930 is sufficiently converged with the desired target value), after which the Al model 930 is considered to be sufficiently trained. The values of the learned parameters are then fixed and the Al model 930 is then deployed to generate output in real-world applications (also referred to as “inference”).

[0094] In some examples, a trained ML model is fine-tuned, meaning that the values of the learned parameters are adjusted slightly in order for the ML model to better model a specific task. Fine-tuning of an Al model 930 typically involves further training the ML model on a number of data samples (which may be smaller in number / cardinality thanthose used to train the model initially) that closely target the specific task. For example, an Al model 930 for generating natural language that has been trained generically on publicly available text corpora is, e.g., fine-tuned by further training using specific training samples. In some embodiments, the specific training samples are used to generate language in a certain style or a certain format. For example, the Al model 930 is trained to generate a blog post having a particular style and structure with a given topic.

[0095] Some concepts in ML-based language models are now discussed. It may be noted that, while the term “language model” has been commonly used to refer to a ML- based language model, there could exist non-ML language models. In the present disclosure, the term “language model” may be used as shorthand for an ML-based language model (i.e., a language model that is implemented using a neural network or other ML architecture), unless stated otherwise. For example, unless stated otherwise, the “language model” encompasses large language models (LLMs) .

[0096] In some embodiments, the language model uses a neural network (typically a DNN) to perform NLP tasks. A language model is trained to model how words relate to each other in a textual sequence, based on probabilities. In some embodiments, the language model contains hundreds of thousands of learned parameters, or in the case of a LLM contains millions or billions of learned parameters or more. As non-limiting examples, a language model can generate text, translate text, summarize text, answer questions, write code (e.g., Python, JavaScript, or other programming languages), classify text (e.g., to identify spam emails), create content for various purposes (e.g., social media content, factual content, or marketing content), or create personalized content for a particular individual or group of individuals. Language models can also be used for chatbots (e.g., virtual assistance).

[0097] In recent years, there has been interest in a type of neural network architecture, referred to as a transformer, for use as language models. For example, the Bidirectional Encoder Representations from Transformers (BERT) model, the Transformer-XL model, and the Generative Pre-trained Transformer (GPT) models are types of transformers. A transformer is a type of neural network architecture that uses self-attention mechanisms in order to generate predicted output based on input data that has some sequential meaning (i.e., the order of the input data is meaningful, which is the case for most text input). Although transformer-based language models are describedherein, it should be understood that the present disclosure may be applicable to any ML- based language model, including language models based on other neural network architectures such as recurrent neural network (RNN)-based language models.

[0098] Although a general transformer architecture for a language model and the model’s theory of operation have been described above, this is not intended to be limiting. Existing language models include language models that are based only on the encoder of the transformer or only on the decoder of the transformer. An encoder-only language model encodes the input text sequence (e.g., device information, entities, electronic artifacts, and so forth) into feature vectors that can then be further processed by a taskspecific layer (e.g., a classification layer). BERT is an example of a language model that is considered to be an encoder-only language model. A decoder-only language model accepts embeddings as input and uses auto-regression to generate an output text sequence. Transformer-XL and GPT-type models are language models that are considered to be decoder-only language models.

[0099] Because GPT-type language models tend to have a large number of parameters, these language models are considered LLMs. An example of a GPT-type LLM is GPT-3. GPT-3 is a type of GPT language model that has been trained (in an unsupervised manner) on a large corpus derived from documents available to the public online. GPT-3 has a very large number of learned parameters (on the order of hundreds of billions), is able to accept a large number of tokens as input (e.g., up to 2,048 input tokens), and is able to generate a large number of tokens as output (e.g., up to 2,048 tokens). GPT-3 has been trained as a generative model, meaning that GPT-3 can process input text sequences to predictively generate a meaningful output text sequence. ChatGPT is built on top of a GPT-type LLM and has been fine-tuned with training datasets based on text-based chats (e.g., chatbot conversations). ChatGPT is designed for processing natural language, receiving chat-like inputs, and generating chat-like outputs.

[0100] A computer system can access a remote language model (e.g., a cloudbased language model), such as ChatGPT or GPT-3, via a software interface (e.g., an API). Additionally or alternatively, such a remote language model can be accessed via a network such as, for example, the Internet. In some embodiments, such as, for example, potentially in the case of a cloud-based language model, a remote language model is hosted by a computer system that includes a plurality of cooperating (e.g., cooperatingvia a network) computer systems that are in, for example, a distributed arrangement. Notably, a remote language model employs a plurality of processors (e.g., hardware processors such as, for example, processors of cooperating computer systems). Indeed, processing of inputs by an LLM can be computationally expensive / can involve a large number of operations (e.g., many instructions can be executed / large data structures can be accessed from memory), and providing output in a required timeframe (e.g., real-time or near real-time) can require the use of a plurality of processors / cooperating computing devices as discussed above.

[0101] In some embodiments, inputs to an LLM are referred to as a prompt (e.g., command set or instruction set), which is a natural language input that includes instructions to the LLM to generate a desired output. In some embodiments, a computer system generates a prompt that is provided as input to the LLM via the LLM’s API. As described above, the prompt is processed or pre-processed into a token sequence prior to being provided as input to the LLM via the LLM’s API. A prompt includes one or more examples of the desired output, which provides the LLM with additional information to enable the LLM to generate output according to the desired output. Additionally or alternatively, the examples included in a prompt provide inputs (e.g., example inputs) corresponding to / as can be expected to result in the desired outputs provided. A one- shot prompt refers to a prompt that includes one example, and a few-shot prompt refers to a prompt that includes multiple examples. A prompt that includes no examples is referred to as a zero-shot prompt.

[0102] In some embodiments, the Ilama2 is used as a large language model, which is a large language model based on an encoder-decoder architecture, and can simultaneously perform text generation and text understanding. The Ilama2 selects or trains proper pre-training corpus, pre-training targets and pre-training parameters according to different tasks and fields, and adjusts a large language model on the basis so as to improve the performance of the large language model under a specific scene.

[0103] In some embodiments, the Falcon40B is used as a large language model, which is a causal decoder-only model. During training, the model predicts the subsequent tokens with a causal language modeling task. The model applies rotational positional embeddings in the model’s transformer model and encodes the absolution positional information of the tokens into a rotation matrix.

[0104] In some embodiments, the Claude is used as a large language model, which is an autoregressive model trained on a large text corpus unsupervised.Computing Platform

[0105] Figure 10 is a block diagram illustrating an example computer system 1000, in accordance with one or more embodiments. In some embodiments, components of the example computer system 1000 are used to implement the software platforms described herein. At least some operations described herein can be implemented on the computer system 1000.

[0106] In some embodiments, the computer system 1000 includes one or more central processing units (“processors”) 1002, main memory 1006, non-volatile memory 1010, network adapters 1012 (e.g., network interface), video displays 1018, input / output devices 1020, control devices 1022 (e.g., keyboard and pointing devices), drive units 1024 including a storage medium 1026, and a signal generation device 1020 that are communicatively connected to a bus 1016. The bus 1016 is illustrated as an abstraction that represents one or more physical buses and / or point-to-point connections that are connected by appropriate bridges, adapters, or controllers. The bus 1016, therefore, includes a system bus, a peripheral component interconnect (PCI) bus or PCI-Express bus, a HyperTransport or industry standard architecture (ISA) bus, a small computer system interface (SCSI) bus, a universal serial bus (USB), IIC (I2C) bus, or an Institute of Electrical and Electronics Engineers (IEEE) standard 1094 bus (also referred to as “Firewire”).

[0107] In some embodiments, the computer system 1000 shares a similar computer processor architecture as that of a desktop computer, tablet computer, personal digital assistant (PDA), mobile phone, game console, music player, wearable electronic device (e.g., a watch or fitness tracker), network-connected (“smart”) device (e.g., a television or home assistant device), virtual / augmented reality systems (e.g., a head-mounted display), or another electronic device capable of executing a set of instructions (sequential or otherwise) that specify action(s) to be taken by the computer system 1000.

[0108] While the main memory 1006, non-volatile memory 1010, and storage medium 1026 (also called a “machine-readable medium”) are shown to be a single medium, the terms “machine-readable medium” and “storage medium” should be taken to include a single medium or multiple media (e.g., a centralized / distributed databaseand / or associated caches and servers) that store one or more sets of instructions 1028. The term “machine-readable medium” and “storage medium” shall also be taken to include any medium that is capable of storing, encoding, or carrying a set of instructions for execution by the computer system 1000. In some embodiments, the non-volatile memory 1010 or the storage medium 1026 is a non-transitory, computer-readable storage medium storing computer instructions, which is executable by one or more “processors” 1002 to perform functions of the embodiments disclosed herein.

[0109] In general, the routines executed to implement the embodiments of the disclosure can be implemented as part of an operating system or a specific application, component, program, object, module, or sequence of instructions (collectively referred to as “computer programs”). The computer programs typically include one or more instructions (e.g., instructions 1004, 1008, 1028) set at various times in various memory and storage devices in a computer device. When read and executed by one or more processors 1002, the instruction(s) cause the computer system 1000 to perform operations to execute elements involving the various aspects of the disclosure.

[0110] Moreover, while embodiments have been described in the context of fully functioning computer devices, those skilled in the art will appreciate that the various embodiments are capable of being distributed as a program product in a variety of forms. The disclosure applies regardless of the particular type of machine or computer-readable media used to actually affect the distribution.

[0111] Further examples of machine-readable storage media, machine-readable media, or computer-readable media include recordable-type media such as volatile and non-volatile memory devices 1010, floppy and other removable disks, hard disk drives, optical discs (e.g., compact disc read-only memory (CD-ROMS), digital versatile discs (DVDs)), and transmission-type media such as digital and analog communication links.

[0112] The network adapter 1012 enables the computer system 1000 to mediate data in a network 1014 with an entity that is external to the computer system 1000 through any communication protocol supported by the computer system 1000 and the external entity. The network adapter 1012 includes a network adapter card, a wireless network interface card, a router, an access point, a wireless router, a switch, a multilayer switch, a protocol converter, a gateway, a bridge, a bridge router, a hub, a digital media receiver, and / or a repeater.

[0113] In some embodiments, the network adapter 1012 includes a firewall that governs and / or manages permission to access proxy data in a computer network and tracks varying levels of trust between different machines and / or applications. The firewall is any number of modules having any combination of hardware and / or software components able to enforce a predetermined set of access rights between a particular set of machines and applications, machines and machines, and / or applications and applications (e.g., to regulate the flow of traffic and resource sharing between these entities). In some embodiments, the firewall additionally manages and / or has access to an access control list that details permissions, including the access and operation rights of an object by an individual, a machine, and / or an application, and the circumstances under which the permission rights stand.

[0114] The techniques introduced here can be implemented by programmable circuitry (e.g., one or more microprocessors), software and / or firmware, special-purpose hardwired (i.e., non-programmable) circuitry, or a combination of such forms. Specialpurpose circuitry can be in the form of one or more application-specific integrated circuits (ASICs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), etc. A portion of the methods described herein can be performed using the example Al system 900 illustrated and described in more detail with reference to Figure 9.Examples

[0115] The present technology is illustrated, for example, according to various aspects described below. Various examples of aspects of the present technology are described as numbered examples (1 , 2, 3, etc.) for convenience. These are provided as examples and do not limit the present technology. It is noted that any of the dependent examples can be combined in any suitable manner, and placed into a respective independent example. The other examples can be presented in a similar manner.1. A computer-implemented method for allocating electronic devices, the method comprising: causing a first display, via a user interface of a computing device, of an inventory identifying a set of electronic devices each associated with device information that includes one or more of: a unique identifier, a name, asource identifier, an associated inventory quantity, a condition, a grade, a carrier, or a model; receiving, via the user interface, a user request representing a set of electronic artifacts that each includes an identification of (i) a subset of electronic devices within the set of electronic devices and (ii) a set of resources configured to be exchanged for the subset of electronic devices, wherein each electronic artifact is linked to an entity associated with a set of attributes that include one or more of: a historical order volume, a geographic location, or a distribution network size; responsive to receiving the user request, transmitting an input including (i) the set of electronic artifacts, (ii) respective device information of each electronic device in the set of electronic devices, and (iii) the set of attributes of the entity linked to each electronic artifact into an artificial intelligence model trained to generate an output including an identification of (i) a subset of electronic artifacts and (ii) allocated electronic devices for each of the subset of electronic artifacts by: determining, for each electronic artifact in the set of electronic artifacts, a ratio between the set of resources and a quantity of electronic devices in the subset of electronic devices, assigning, for each electronic device in the set of electronic devices, a weight to the entity linked to each electronic artifact based on (i) the device information of each electronic device and (ii) the set of attributes of the entity, and allocating each electronic device in the set of electronic devices to one or more electronic artifacts in accordance with (i) respective ratios of the one or more electronic artifacts and (ii) respective assigned weight of the entity linked to the one or more electronic artifacts; causing a second display, via the user interface, of a representation of (i) the subset of electronic artifacts and (ii) the allocated electronic devices for each of the subset of electronic artifacts; and transmitting, to one or more entities associated with the subset of electronic artifacts, a notification indicating acceptance of the subset of electronic artifacts.2. The computer-implemented method of Example 1 , wherein the artificial intelligence model is trained using historical data including one or more of: historical allocation of historical sets of electronic devices or one or more patterns associated with the linked entities.3. The computer-implemented method of Example 1 , wherein the user input is a first user input, wherein the set of electronic artifacts is a first set of electronic artifacts, wherein the subset of electronic devices is a first subset of electronic devices, wherein the set of resources is a first set of resources, further comprising: receiving, via the user interface, a second user request representing a second set of electronic artifacts that each includes an identification of (i) a second subset of electronic devices within the set of electronic devices and (ii) a second set of resources configured to be exchanged for the second subset of electronic devices, wherein the second set of resources is larger than the first set of resources; and responsive to receiving the second user input, inputting (i) the second set of electronic artifacts, (ii) respective device information of each electronic device in the second set of electronic devices, and (iii) a respective set of attributes of a respective entity linked to each of the second set of electronic artifacts into the artificial intelligence model trained to output an identification of (i) a second subset of electronic artifacts and (ii) allocated electronic devices for each of the second subset of electronic artifacts.4. The computer-implemented method of Example 1 , wherein the user input indicates a maximum resource allocation limit for a particular entity associated with multiple electronic artifacts, further comprising: tracking a cumulative resource allocation for the particular entity across the multiple electronic artifacts; comparing the cumulative resource allocation to the maximum resource allocation limit; and responsive to the cumulative resource allocation exceeding the maximum resource allocation limit, preventing allocation of subsequently allocated electronic devices of the set of electronic devices to the particular entity.5. The computer-implemented method of Example 1 , wherein the artificial intelligence model includes a neural network with an encoder-decoder architecture trained to: encode the user request into a first set of latent space representations; compare the first set of latent space representations with a second set of latent space representations derived from the set of electronic devices to determine the allocated electronic devices for each of the subset of electronic artifacts; and decode respective latent space representations of the subset of electronic artifacts to generate the output.6. The computer-implemented method of Example 1 , further comprising: determine that a first electronic artifact and a second electronic artifact both specify a common electronic device; comparing respective sets of resources of the first electronic artifact and the second electronic artifact; and allocating the common electronic device to the electronic artifact having a larger set of resources.7. The computer-implemented method of Example 1 , further comprising: determining that a first electronic artifact and a second electronic artifact have equal sets of resources; comparing a first maximum quantity of the first electronic artifact to a second maximum quantity of the second electronic artifact; and allocating an associated electronic device to the electronic artifact having a higher maximum quantity.8. The computer-implemented method of Example 1 , wherein allocating each electronic device comprises, for each electronic device identified by the inventory: ranking the set of electronic artifacts based on respective ratios and assigned weights; and allocating the electronic device to a particular electronic artifact having a highest rank.9. The computer-implemented method of Example 1 , wherein the output represents a set of allocation recommendations for the set of electronic artifacts, further comprising: receiving, via the user interface, a user selection approving one or more allocation recommendations from the output; and initiating one or more actions to distribute the allocated electronic devices of each of the subset of electronic artifacts to respective entities.10. The computer-implemented method of Example 1 , further comprising: generating a numerical score for each attribute in the set of attributes; and aggregating the numerical score for each attribute in the set of attributes to determine the weight.11. A non-transitory, computer-readable storage medium comprising instructions thereon, wherein the instructions, when executed by at least one data processor of a system, cause the system to: cause a first display, via a user interface associated with an entity, of an inventory identifying a set of electronic devices each associated with device information, wherein the entity is linked to a set of attributes; receive, via the user interface, a user request representing a set of electronic artifacts that each includes an identification of (i) one or more electronic devices within the set of electronic devices and (ii) a set of resources configured to be exchanged for the one or more electronic devices; determine (i) an acceptance status of one or more electronic artifacts within the set of electronic artifacts and (ii) allocated electronic devices corresponding to the set of electronic artifacts based on (i) a ratio between the set of resources and a quantity of electronic devices in the one or more electronic devices and (ii) a weight assigned to the entity, wherein the weight is determined based on (i) the device information of each electronic device and (ii) the set of attributes of the entity; and cause a second display of, via the user interface, the representation of the acceptance status and the allocated electronic devices.12. The non-transitory, computer-readable storage medium of Example 11 , wherein the subset of electronic artifacts is generated using an artificial intelligence model.13. The non-transitory, computer-readable storage medium of Example 11 , wherein the system is further caused to: receive, via the user interface, a set of allocation constraints including one or more of: a maximum quantity of electronic devices per entity, a minimum quantity of electronic devices per entity, or a device type restriction per entity, wherein allocating each electronic device is in accordance with the set of allocation constraints.14. The non-transitory, computer-readable storage medium of Example 11 , wherein the system is further caused to: cause initiation of a transaction for each electronic artifact in the subset of electronic artifacts based on the subset of electronic artifacts.15. The non-transitory, computer-readable storage medium of Example 11 , wherein the system is further caused to: cause generation of a report indicating the subset of electronic artifacts, wherein the report includes a description indicating a respective ratio and assigned weight for each allocation.16. A system for allocating electronic devices, the system comprising: at least one hardware processor; and at least one non-transitory memory storing instructions, which, when executed by the at least one hardware processor, cause the system to: cause a first display, via a user interface of a computing device, of an inventory identifying a set of electronic devices each associated with device information; receive, via the user interface, a user request representing a set of electronic artifacts that each includes an identification of (i) one or more electronic devices within the set of electronic devices and (ii)a set of resources configured to be exchanged for the one or more electronic devices, wherein each electronic artifact is linked to an entity associated with a set of attributes; generate a subset of electronic artifacts by allocating each electronic device in the set of electronic devices to one or more electronic artifacts in accordance with (i) a ratio between the set of resources and a quantity of electronic devices in the one or more electronic devices and (ii) a weight to the entity linked to each electronic artifact, wherein the weight is determined based on (i) the device information of each electronic device and (ii) the set of attributes of the entity; and cause a second display of, via the user interface, a representation of the subset of electronic artifacts.17. The system of Example 16, wherein the device information of the set of electronic devices are stored in one or more of: a cloud environment hosted by a cloud provider or a self-hosted environment.18. The system of Example 16, wherein the system is further caused to: receive, via the user interface, a first set of electronic artifacts during a first bidding round; and receive, via the user interface, a second set of electronic artifacts during a second bidding round, wherein the set of electronic artifacts comprises the first set of electronic artifacts and the second set of electronic artifacts.19. The system of Example 18, wherein the second bidding round occurs after the first bidding round, and wherein electronic artifacts received during the second bidding round are restricted to having a larger set of resources than corresponding electronic artifacts received during the first bidding round.20. The system of Example 16, wherein the set of artifacts is a first set of electronic artifacts, and wherein the system is further caused to: determine a first allocation of electronic devices based on the first set of electronic artifacts; cause a third display, via the user interface, of the first allocation; receive, via the user interface, a second set of electronic artifacts; and determine a second allocation of electronic devices based on the first set of electronic artifacts and the second set of electronic artifacts.Conclusion

[0116] Unless the context clearly requires otherwise, throughout the description and the claims, the words “comprise,” “comprising,” and the like are to be construed in an inclusive sense, as opposed to an exclusive or exhaustive sense; that is to say, in the sense of “including, but not limited to.” As used herein, the terms “connected,” “coupled,” or any variant thereof means any connection or coupling, either direct or indirect, between two or more elements; the coupling or connection between the elements can be physical, logical, or a combination thereof. Additionally, the words “herein,” “above,” “below,” and words of similar import, when used in this application, refer to this application as a whole and not to any particular portions of this application. Where the context permits, words in the above Detailed Description using the singular or plural number may also include the plural or singular number respectively. The word “or,” in reference to a list of two or more items, covers all of the following interpretations of the word: any of the items in the list, all of the items in the list, and any combination of the items in the list.

[0117] The above Detailed Description of examples of the technology is not intended to be exhaustive or to limit the technology to the precise form disclosed above. While specific examples for the technology are described above for illustrative purposes, various equivalent modifications are possible within the scope of the technology, as those skilled in the relevant art will recognize. For example, while processes or blocks are presented in a given order, alternative embodiments may perform routines having steps, or employ systems having blocks, in a different order, and some processes or blocks may be deleted, moved, added, subdivided, combined, and / or modified to provide alternative or sub-combinations. Each of these processes or blocks may be implemented in a variety of different ways. Also, while processes or blocks are at times shown as beingperformed in series, these processes or blocks may instead be performed or implemented in parallel, or may be performed at different times. Further, any specific numbers noted herein are only examples: alternative embodiments may employ differing values or ranges.

[0118] The teachings of the technology provided herein can be applied to other systems, not necessarily the system described above. The elements and acts of the various examples described above can be combined to provide further embodiments of the technology. Some alternative embodiments of the technology may include not only additional elements to those embodiments noted above, but also may include fewer elements.

[0119] These and other changes can be made to the technology in light of the above Detailed Description. While the above description describes certain examples of the technology, and describes the best mode contemplated, no matter how detailed the above appears in text, the technology can be practiced in many ways. Details of the system may vary considerably in its specific embodiment, while still being encompassed by the technology disclosed herein. As noted above, specific terminology used when describing certain features or aspects of the technology should not be taken to imply that the terminology is being redefined herein to be restricted to any specific characteristics, features, or aspects of the technology with which that terminology is associated. In general, the terms used in the following claims should not be construed to limit the technology to the specific examples disclosed in the specification, unless the above Detailed Description section explicitly defines such terms. Accordingly, the actual scope of the technology encompasses not only the disclosed examples, but also all equivalent ways of practicing or implementing the technology under the claims.

[0120] Certain aspects of the technology are presented below in certain claim forms, but the applicant contemplates the various aspects of the technology in any number of claim forms. For example, while some aspects of the technology may be recited as a computer-readable medium claim, other aspects may likewise be embodied as a computer-readable medium claim, or in other forms, such as being embodied in a means-plus-function claim. Accordingly, the applicant reserves the right to pursue additional claims after filing this application to pursue such additional claim forms, in either this application or in a continuing application.

Claims

CLAIMSWe claim:

1. A computer-implemented method for allocating electronic devices, the method comprising: causing a first display, via a user interface of a computing device, of an inventory identifying a set of electronic devices each associated with device information that includes one or more of: a unique identifier, a name, a source identifier, an associated inventory quantity, a condition, a grade, a carrier, or a model; receiving, via the user interface, a user request representing a set of electronic artifacts that each includes an identification of (i) a subset of electronic devices within the set of electronic devices and (ii) a set of resources configured to be exchanged for the subset of electronic devices, wherein each electronic artifact is linked to an entity associated with a set of attributes that include one or more of: a historical order volume, a geographic location, or a distribution network size; responsive to receiving the user request, transmitting an input including (i) the set of electronic artifacts, (ii) respective device information of each electronic device in the set of electronic devices, and (iii) the set of attributes of the entity linked to each electronic artifact into an artificial intelligence model trained to generate an output including an identification of (i) a subset of electronic artifacts and (ii) allocated electronic devices for each of the subset of electronic artifacts by: determining, for each electronic artifact in the set of electronic artifacts, a ratio between the set of resources and a quantity of electronic devices in the subset of electronic devices, assigning, for each electronic device in the set of electronic devices, a weight to the entity linked to each electronic artifact based on (i) the device information of each electronic device and (ii) the set of attributes of the entity, andallocating each electronic device in the set of electronic devices to one or more electronic artifacts in accordance with (i) respective ratios of the one or more electronic artifacts and (ii) respective assigned weight of the entity linked to the one or more electronic artifacts; causing a second display, via the user interface, of a representation of (i) the subset of electronic artifacts and (ii) the allocated electronic devices for each of the subset of electronic artifacts; and transmitting, to one or more entities associated with the subset of electronic artifacts, a notification indicating acceptance of the subset of electronic artifacts.

2. The computer-implemented method of claim 1 , wherein the artificial intelligence model is trained using historical data including one or more of: historical allocation of historical sets of electronic devices or one or more patterns associated with the linked entities.

3. The computer-implemented method of claim 1 , wherein the user input is a first user input, wherein the set of electronic artifacts is a first set of electronic artifacts, wherein the subset of electronic devices is a first subset of electronic devices, wherein the set of resources is a first set of resources, further comprising: receiving, via the user interface, a second user request representing a second set of electronic artifacts that each includes an identification of (i) a second subset of electronic devices within the set of electronic devices and (ii) a second set of resources configured to be exchanged for the second subset of electronic devices, wherein the second set of resources is larger than the first set of resources; and responsive to receiving the second user input, inputting (i) the second set of electronic artifacts, (ii) respective device information of each electronic device in the second set of electronic devices, and (iii) a respective set of attributes of a respective entity linked to each of the second set of electronic artifacts into the artificial intelligence model trained to output an identification of (i) a second subset of electronic artifacts and (ii) allocated electronic devices for each of the second subset of electronic artifacts.

4. The computer-implemented method of claim 1 , wherein the user input indicates a maximum resource allocation limit for a particular entity associated with multiple electronic artifacts, further comprising: tracking a cumulative resource allocation for the particular entity across the multiple electronic artifacts; comparing the cumulative resource allocation to the maximum resource allocation limit; and responsive to the cumulative resource allocation exceeding the maximum resource allocation limit, preventing allocation of subsequently allocated electronic devices of the set of electronic devices to the particular entity.

5. The computer-implemented method of claim 1 , wherein the artificial intelligence model includes a neural network with an encoder-decoder architecture trained to: encode the user request into a first set of latent space representations; compare the first set of latent space representations with a second set of latent space representations derived from the set of electronic devices to determine the allocated electronic devices for each of the subset of electronic artifacts; and decode respective latent space representations of the subset of electronic artifacts to generate the output.

6. The computer-implemented method of claim 1 , further comprising: determine that a first electronic artifact and a second electronic artifact both specify a common electronic device; comparing respective sets of resources of the first electronic artifact and the second electronic artifact; and allocating the common electronic device to the electronic artifact having a larger set of resources.

7. The computer-implemented method of claim 1 , further comprising: determining that a first electronic artifact and a second electronic artifact have equal sets of resources;comparing a first maximum quantity of the first electronic artifact to a second maximum quantity of the second electronic artifact; and allocating an associated electronic device to the electronic artifact having a higher maximum quantity.

8. The computer-implemented method of claim 1 , wherein allocating each electronic device comprises, for each electronic device identified by the inventory: ranking the set of electronic artifacts based on respective ratios and assigned weights; and allocating the electronic device to a particular electronic artifact having a highest rank.

9. The computer-implemented method of claim 1 , wherein the output represents a set of allocation recommendations for the set of electronic artifacts, further comprising: receiving, via the user interface, a user selection approving one or more allocation recommendations from the output; and initiating one or more actions to distribute the allocated electronic devices of each of the subset of electronic artifacts to respective entities.

10. The computer-implemented method of claim 1 , further comprising: generating a numerical score for each attribute in the set of attributes; and aggregating the numerical score for each attribute in the set of attributes to determine the weight.

11. A non-transitory, computer-readable storage medium comprising instructions thereon, wherein the instructions, when executed by at least one data processor of a system, cause the system to: cause a first display, via a user interface associated with an entity, of an inventory identifying a set of electronic devices each associated with device information, wherein the entity is linked to a set of attributes; receive, via the user interface, a user request representing a set of electronic artifacts that each includes an identification of (i) one or more electronicdevices within the set of electronic devices and (ii) a set of resources configured to be exchanged for the one or more electronic devices; determine (i) an acceptance status of one or more electronic artifacts within the set of electronic artifacts and (ii) allocated electronic devices corresponding to the set of electronic artifacts based on (i) a ratio between the set of resources and a quantity of electronic devices in the one or more electronic devices and (ii) a weight assigned to the entity, wherein the weight is determined based on (i) the device information of each electronic device and (ii) the set of attributes of the entity; and cause a second display of, via the user interface, the representation of the acceptance status and the allocated electronic devices.

12. The non-transitory, computer-readable storage medium of claim 11 , wherein the subset of electronic artifacts is generated using an artificial intelligence model.

13. The non-transitory, computer-readable storage medium of claim 11 , wherein the system is further caused to: receive, via the user interface, an identification of a set of allocation constraints including one or more of: a maximum quantity of electronic devices per entity, a minimum quantity of electronic devices per entity, or a device type restriction per entity, wherein allocating each electronic device is in accordance with the set of allocation constraints.

14. The non-transitory, computer-readable storage medium of claim 11 , wherein the system is further caused to: cause initiation of a transaction for each electronic artifact in the subset of electronic artifacts based on the subset of electronic artifacts.

15. The non-transitory, computer-readable storage medium of claim 11 , wherein the system is further caused to:cause generation of a report indicating the subset of electronic artifacts, wherein the report includes a description indicating a respective ratio and assigned weight for each allocation.

16. A system for allocating electronic devices, the system comprising: at least one hardware processor; and at least one non-transitory memory storing instructions, which, when executed by the at least one hardware processor, cause the system to: cause a first display, via a user interface of a computing device, of an inventory identifying a set of electronic devices each associated with device information; receive, via the user interface, a user request representing a set of electronic artifacts that each includes an identification of (i) one or more electronic devices within the set of electronic devices and (ii) a set of resources configured to be exchanged for the one or more electronic devices, wherein each electronic artifact is linked to an entity associated with a set of attributes; generate a subset of electronic artifacts by allocating each electronic device in the set of electronic devices to one or more electronic artifacts in accordance with (i) a ratio between the set of resources and a quantity of electronic devices in the one or more electronic devices and (ii) a weight to the entity linked to each electronic artifact, wherein the weight is determined based on (i) the device information of each electronic device and (ii) the set of attributes of the entity; and cause a second display of, via the user interface, a representation of the subset of electronic artifacts.

17. The system of claim 16, wherein the device information of the set of electronic devices are stored in one or more of: a cloud environment hosted by a cloud provider or a self-hosted environment.

18. The system of claim 16, wherein the system is further caused to: receive, via the user interface, a first set of electronic artifacts during a first bidding round; and receive, via the user interface, a second set of electronic artifacts during a second bidding round, wherein the set of electronic artifacts comprises the first set of electronic artifacts and the second set of electronic artifacts.

19. The system of claim 18, wherein the second bidding round occurs after the first bidding round, and wherein electronic artifacts received during the second bidding round are restricted to having a larger set of resources than corresponding electronic artifacts received during the first bidding round.

20. The system of claim 16, wherein the set of artifacts is a first set of electronic artifacts, and wherein the system is further caused to: determine a first allocation of electronic devices based on the first set of electronic artifacts; cause a third display, via the user interface, of the first allocation; receive, via the user interface, a second set of electronic artifacts; and determine a second allocation of electronic devices based on the first set of electronic artifacts and the second set of electronic artifacts.

Images