Instant messaging group cooperation method and device based on intelligent agent, medium and program product

Abstract

The invention discloses an agent-based instant messaging group cooperation method and device, a medium and a program product, and is applied to a client, and the method comprises the steps: sending a super group creation request to an instant messaging server, and triggering the instant messaging server to automatically bind a default digital intelligent employee for a super group as a super group assistant; when a dialogue message is generated in the group, pushing the dialogue message to the instant messaging server, and triggering the instant messaging server to send message data to the intelligent server; triggering the intelligent server to carry out real-time intention analysis, and assembling service message data when a service intention is recognized; the service message data from the intelligent server is received through the instant messaging server, replied in a super group assistant identity and displayed in a service message card form, and a corresponding service process can be triggered through clicking. According to the invention, the AI intelligent agent and the instant messaging group are deeply combined, the communication-service-intelligence fused cooperation is realized by using group communication, and the enterprise cooperation efficiency and the service processing capability are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of artificial intelligence and enterprise collaboration technology, in particular to an instant messaging group collaboration method, device, medium and program product based on an agent. BACKGROUND

[0002] At present, the instant messaging tools and collaboration technologies in the toB (enterprise-oriented) scenario mainly present the following characteristics:

[0003] 1. Limited functions of instant messaging tools:

[0004] The core functions of existing enterprise-level IM tools still revolve around "message sending and receiving, file transmission, meeting reservation", although some tools support "robot plug-ins" (such as automatic reply robots, reminder robots), but such robots only have "fixed instruction response" capabilities (such as keyword triggering preset replies, sending fixed reminders at regular intervals), and cannot understand the context semantics of group conversations, nor can they intelligently summarize conversation content, identify business needs, and dispatch cross-system business.

[0005] 2. Separation of business systems and communication scenarios:

[0006] The core business of an enterprise (such as project task allocation, customer demand follow-up, order processing) usually relies on independent business systems, and business initiation and collaboration need to be communicated and confirmed in the IM group, and then manually jump to the business system for operation — this process has a "communication-business" connection gap, which easily leads to business response lag (such as confirming customer demand in the group, manually logging into the CRM system to create a follow-up task), operation redundancy (repeatedly entering information), and cannot automatically associate business data based on group communication content.

[0007] 3. Inefficiency of group information management and knowledge sedimentation:

[0008] In enterprise group collaboration, as the conversation time increases, the messages in the group present the characteristics of "fragmentation, multi-threading": new people entering the group need to manually review historical messages to understand the collaboration background, which is time-consuming and easy to miss key information; the business conclusions and to-do items discussed in the group need to be manually organized and archived, otherwise they are easily covered by new messages; when a user raises a business question, they need to manually search group files, enterprise knowledge bases, or ask old employees for answers, and cannot immediately get accurate answers — the above problems lead to low reuse rate of group information and high cost of knowledge sedimentation.

[0009] 4. Limited application of AI technology in enterprise collaboration:

[0010] In the prior art, AI large models such as ChatGPT and enterprise private large models have the ability of text summarization and knowledge question and answer, but they are not deeply integrated with enterprise IM groups. On the one hand, the large model cannot obtain group conversation data and group file information in real time, resulting in that the summary or answer generated by the large model is out of the context of group collaboration. On the other hand, the large model lacks the ability to interface with enterprise business systems, and cannot convert “business demand identification” into “business portal triggering” or “business intelligent agent awakening”, and cannot realize the closed loop of “intelligence - business”.

[0011] In view of the above status quo, in the toB enterprise instant messaging collaboration scene, the prior art mainly has the following defects that cannot be ignored:

[0012] The new person group entry guidance lacks intelligence and the collaboration threshold is high. In the prior art, the new person group entry needs to rely on manual sending of welcome speech and manual sorting of historical conversation core content, which not only consumes the time cost of old employees, but also easily causes the new person to be unable to quickly integrate into collaboration due to incomplete manual summary. Especially when the group historical conversation exceeds one thousand, the efficiency and accuracy of manual sorting are greatly reduced, and the adaptation cycle of the new person is prolonged.

[0013] The group internal to-do list and problem response lack automatic reminding, and key collaboration nodes are easily missed. The prior art cannot identify the “to-be-handled problem” (such as “XX responsible for the project scheme needs to be fed back by Friday”) in the group in real time, and cannot trigger automatic reminding for the case that “the responsible person has not responded for a long time”, resulting in delay of collaboration nodes (such as customer demand feedback overtime, cross-department collaboration task shelving), affecting the overall business promotion efficiency.

[0014] The group internal conversation conclusion and business demand lack intelligent conversion, and the communication and business are disconnected. The prior art can only realize “message transmission”, and cannot identify business demand based on group internal conversation content (such as “we need to have a meeting at 3 pm today to discuss the problem of customer demand”), and cannot automatically trigger business portal or awaken business intelligent agent (such as schedule intelligent agent) — human intervention is needed to jump to the business system operation, increasing the business processing steps, and the business information input is easily wrong due to human memory bias.

[0015] The business knowledge question and answer lack precision and context association, and the knowledge reuse rate is low. In the prior art, when a user raises a business question (such as “what is the historical service agreement content of A customer”) in the group, manual retrieval of group files, enterprise knowledge base or consultation of relevant personnel is needed: manual retrieval is time-consuming and easy to miss key information, and general AI question and answer tools cannot access group private data (group messages, group files) and enterprise knowledge base, resulting in that the answer generated lacks pertinence and cannot meet the precise question and answer demand in the enterprise business scene, leading to that the knowledge assets cannot be efficiently reused.

[0016] The lack of an integrated architecture for "communication-business-intelligence" leads to significant bottlenecks in collaboration efficiency. Currently, instant messaging tools, business systems, and AI technologies operate independently: communication data cannot be synchronized to business systems and AI models; the status of business systems cannot be fed back to the communication context; and the output of AI models cannot directly trigger business operations—creating "data silos" and "business disconnects." This results in enterprise collaboration efficiency being limited by "manual coordination," preventing the automation and intelligent upgrading of collaboration processes through AI technology.

[0017] In summary, existing technologies cannot meet the collaborative needs of B2B enterprises in integrating "communication, business, and intelligence." There is an urgent need for a technical solution that can deeply integrate AI agents with instant messaging groups to address the aforementioned shortcomings and improve enterprise collaboration efficiency and business processing capabilities. Summary of the Invention

[0018] In view of this, the present invention provides a method, device, medium and program product for real-time communication group collaboration based on intelligent agents, which at least partially solves the problems existing in the prior art.

[0019] To achieve the above objectives, the present invention adopts the following technical solution:

[0020] In a first aspect, the present invention provides an agent-based instant messaging group collaboration method applied to a client, the method comprising:

[0021] Send a supergroup creation request to the instant messaging server, triggering the instant messaging server to automatically bind a default digital employee as the supergroup assistant;

[0022] When a conversation occurs in the super group, the message is pushed to the instant messaging server, which then sends the message data to the intelligent server. The intelligent server is then triggered to perform real-time intent analysis through the intent analysis process to determine whether the message contains business intent. When the business intent is identified, it is assembled into business message data.

[0023] The system receives business message data from the intelligent server via an instant messaging server, replies as a super group assistant, and displays the message in the form of a business message card. The business message card is set to trigger the corresponding business process by clicking on it.

[0024] As a further improvement of the present invention, the step of triggering the instant messaging server to send message data to the intelligent server includes:

[0025] The instant messaging server is triggered to synchronously push message data to the search engine database, which in turn triggers the search engine database to store the message data, convert the message into an intent analysis task, and send the intent analysis task to the intelligent server via synchronous push or asynchronous scheduling.

[0026] Furthermore, digital employees are tagged with exclusive fields and have their own UI interaction rules, address book permissions, and business capability scope defined.

[0027] In addition to the default digital employees that are bound to the system, other business-related digital employees can be added as needed through the contact list selection component, within the scope of the enterprise's digital employee permissions.

[0028] The instant messaging server processes the sending and receiving of messages for digital employees based on user ID.

[0029] Furthermore, the business message card includes a button component, which defines an action event type and supports:

[0030] openUrl: Opens the business H5 page;

[0031] openApp: Redirects to the specified business page within the enterprise app;

[0032] openAgent: Redirects to the independent interactive page for digital employees;

[0033] requestApi: Directly calls the business interface to perform operations;

[0034] When a user clicks a button in a business message card, the action type is parsed, and the corresponding redirect or API call is triggered.

[0035] Furthermore, including:

[0036] When the business system status changes, the client receives the updated business message JSON received through the instant messaging server's message modification interface when the business system status changes, and displays the updated status on the business message card.

[0037] Furthermore, the business messages received by the client are processed using a unified cross-client business message rendering method, including: using a unified business message rendering template, defining component types and exclusive attributes in JSON format, and adding common control fields;

[0038] After receiving business message data, each client parses the JSON template and renders the UI according to its own technology stack to ensure consistent display style across different clients.

[0039] Furthermore, it also includes a scheduled automation service, which includes:

[0040] When a super group creation request is sent to the instant messaging server, the instant messaging server is triggered to automatically register a periodic scheduled task for the super group assistant and store the task information in the scheduled task scheduling center.

[0041] After the scheduled task is triggered, it requests the scheduling center to send the task to the intelligent server via synchronous push or asynchronous scheduling.

[0042] And trigger the intelligent server to call the scheduled task intent stream for contextual message processing;

[0043] Receive contextualized message processing results from the intelligent server via the instant messaging server, and reply as a super group assistant.

[0044] Furthermore, the scenario-based message processing includes: determining the processing scenario of the scheduled task according to the super group's preset skill strategy; filtering group messages within a specified time range according to scenario requirements; and for each message group, calling the large model through a dedicated Prompt to determine whether it matches the scenario, and executing the corresponding operation if it matches.

[0045] The scenario requirements include handling user non-response to @ messages, summarizing periodic conversations, or handling tasks assigned within a group.

[0046] Furthermore, it also includes group-specific intelligent Q&A, which includes:

[0047] When a message @a digital employee is received in the supergroup, the message is synchronized to the intelligent server via the instant messaging server. The intelligent server is then requested to parse the message, trigger an intent stream based on the parsing result, and obtain the processing result by executing the intent stream. The intelligent server's parsing of the message, triggering the intent stream based on the parsing result, and obtaining the processing result by executing the intent stream includes: parsing message tags, extracting message context, and triggering a dedicated question-and-answer intent stream; the dedicated question-and-answer intent stream includes: filtering meaningless content; removing invalid text other than questions; calling the RAG service to customize the recall scope according to the supergroup scenario, the recall scope including: the search engine database corresponding to the supergroup, the enterprise general knowledge base, and / or the FAQ database; inputting multiple results recalled by RAG into a large model to generate a concise and unified answer; if RAG has no matching results, calling the large model to generate a basic answer.

[0048] Receive processing results from the intelligent server via the instant messaging server and receive a reply from the user who was tagged as a digital intelligence employee.

[0049] In a second aspect, the present invention also provides a computer device, the device comprising: a processor and a memory;

[0050] The memory is used to store one or more program instructions;

[0051] The processor is configured to run one or more program instructions to perform the steps of the agent-based instant messaging group collaboration method described above.

[0052] Thirdly, the present invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the agent-based instant messaging group collaboration method described above.

[0053] Fourthly, the present invention also provides a computer program product comprising computer program instructions that, when executed by a processor, implement the steps of the agent-based instant messaging group collaboration method described above.

[0054] This invention achieves a closed loop of "communication-business-intelligence" within enterprises by integrating "supergroup + intelligent agent". It solves problems such as delayed business response, fragmented information, low knowledge reuse rate and high cost of manual collaboration in enterprise collaboration, improves team collaboration efficiency and business processing accuracy, and helps enterprises upgrade their collaboration mode in the process of digital transformation. Attached Figure Description

[0055] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings in the following description are merely exemplary, and those skilled in the art can derive other embodiments based on the provided drawings without creative effort.

[0056] Figure 1 This is a timing diagram of real-time message analysis according to an embodiment of the present invention;

[0057] Figure 2 This is a flowchart of a timed service processing according to an embodiment of the present invention;

[0058] Figure 3 This is a flowchart of an agent-based instant messaging group collaboration method according to an embodiment of the present invention (applied to a client). Detailed Implementation

[0059] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0060] With the deepening of enterprise digital transformation and the rise of intelligence, instant messaging (IM) tools have become the core carrier of internal enterprise collaboration. However, their functions are still limited to "message transmission" and their integration with enterprise business systems (such as project management systems, customer relationship management systems, and enterprise knowledge bases) and artificial intelligence technologies is extremely low. This leads to a disconnect between "communication" and "business" and "intelligence", forming a collaboration barrier.

[0061] Analysis revealed the following issues that require further investigation:

[0062] 1. Identity adaptation issue between intelligent agents and supergroups: Existing enterprise instant messaging supergroup systems only support the "ordinary employee" identity system and lack exclusive identity definition and access control mechanism for intelligent agents. As a result, intelligent agents cannot enter the supergroup with a legitimate and distinguishable identity, nor can their permission scope, interaction mode and functional positioning (such as "group management" "business response" and other sub-roles) be clearly defined in the group.

[0063] 2. Lack of efficiency improvement mechanisms in super group conversations: Existing super groups only provide basic message sending and receiving functions and lack efficiency optimization capabilities based on intelligent analysis. On the one hand, they cannot identify users' business intentions (such as "create schedule" or "initiate approval") in real time and proactively trigger business processes. On the other hand, they cannot provide automated services such as timed reminders and content summaries based on group interaction scenarios (such as "user does not respond to @ message" or "periodic dialogue summary"). This leads to a disconnect between group communication and business processing, resulting in low efficiency.

[0064] 3. The problem of seamless communication and business integration: The existing system suffers from data silos between communication and business: First, there is a lack of a unified business message rendering standard across clients (Android, iOS, HarmonyOS, H5), resulting in inconsistent display styles of complex business messages (such as business cards with buttons and charts) on different platforms; second, IM messages are "one-time immutable" data, which cannot synchronize changes in business status (such as "schedule time modification" and "approval status update"), causing business information to lag within the group; third, there is a lack of direct jump capability from business messages to business processing steps, requiring users to manually operate the business system outside the super group, resulting in a broken workflow.

[0065] 4. The problem of knowledge transformation of communication content in supergroups: Communication dialogues (text, file association information, etc.) within supergroups are only stored in the IM server database, which are "unstructured communication data" and cannot be perceived, retrieved, or reused by intelligent agents; existing IM services, as underlying communication components, do not have the ability to transform communication content into "structured enterprise knowledge data", resulting in the inability of accumulated business information within the group to form reusable knowledge assets.

[0066] 5. Precise question answering by intelligent agents based on enterprise knowledge: Most existing intelligent agents operate on a "one-to-one independent interaction" model, which cannot adapt to the "many-to-many group conversation" scenario of supergroups. They lack both a recognition mechanism for group interaction commands such as "user@intelligent agent" and a precise answering process that combines supergroup-specific knowledge (group history dialogues, group-related files) with enterprise general knowledge (enterprise knowledge base, FAQ), resulting in intelligent agents being unable to respond to business issues efficiently within the group.

[0067] Based on the aforementioned technical problems to be solved, the core objective of this invention is to address the pain points of collaboration in instant messaging groups within B2B enterprise scenarios by deeply integrating AI agents (possessing natural language understanding, knowledge retrieval, and business scheduling capabilities) with group conversations to construct a "super group" system. Through the built-in "super group assistant" agent, it enables real-time response to business needs within the group, intelligent message summarization, automated new member guidance, and precise knowledge Q&A, thereby bridging the data and business flows of "communication-business-intelligence," reducing enterprise collaboration costs, and improving business processing efficiency and knowledge reuse capabilities.

[0068] Specifically, the aforementioned agent-based instant messaging group collaboration method can be applied to intelligent collaboration and business management in B2B (business-to-business) scenarios. It is particularly suitable for enterprise teams that require high-frequency communication, business collaboration, and knowledge accumulation, including but not limited to project teams of Internet companies, production coordination departments of traditional manufacturing industries, customer service teams of financial institutions, and project collaboration groups of consulting firms.

[0069] This embodiment provides an agent-based instant messaging group collaboration method, the overall solution of which is as follows:

[0070] I. Constructing a "Digital Employee" identity system to solve the identity adaptation problem between intelligent agents and supergroups:

[0071] 1. Enterprise Account System Expansion: Add a "Digital Employee" identity type to the existing enterprise account system to structurally differentiate it from the "Regular Employee" identity. This is achieved by marking it with a dedicated field (such as identity_type: "digital_employee") and defining exclusive UI interaction rules (such as special avatar identifiers and conversation bubble styles), address book permissions (only displayed in the enterprise's internal address book, supporting categorization by business dimension), and business capability scope (such as "Finance Digital Employee" only having access to financial domain function interfaces).

[0072] 2. Super swarm intelligence agent admission process:

[0073] Step 1: When a super group is created, the system automatically binds a default "Super Group Assistant" digital employee to the group, who acts as the "chief steward" of the group and is responsible for improving overall communication efficiency and business scheduling;

[0074] Step 2: The group administrator can add business-related digital employees (such as "financial digital employees" and "leave and attendance digital employees") as needed within the enterprise's digital employee permissions through the "Contacts Selection Component".

[0075] Step 3: The IM service only needs to process the sending and receiving of messages by "user ID" and does not need to know the "digital employee" identity attribute, thus reducing the complexity of the underlying service transformation.

[0076] II. Build a knowledge base service to solve the problem of knowledge transformation of supergroup communication content (in conjunction with...) Figure 1 (As shown)

[0077] 1. Deployment of independent knowledge base service: Deploy a knowledge base service that supports massive data retrieval, using Elasticsearch (ES) as the storage engine, specifically for receiving and structuring communication content within the supergroup;

[0078] 2. Communication Content Synchronization Mechanism: When valid dialogue content (such as text messages or key information summaries after file parsing) is generated in the super group, the IM service triggers the synchronization interface of the knowledge base service to push the message data (including message ID, sending time, sender, content text, associated file ID, etc.) to the knowledge base service;

[0079] 3. Content structuring: After receiving data, the knowledge base service automatically performs structuring processing (such as extracting keywords and dividing business domain tags) and stores it in Elasticsearch to form "structured enterprise knowledge data" that can be retrieved by intelligent agents, realizing the transformation of "communication content → knowledge assets".

[0080] Third, design a group-specific intelligent question-and-answer process to solve the problem of accurate question-and-answering by intelligent agents based on enterprise knowledge.

[0081] 1. Group interaction command recognition: Extend the IM message entity object to add the @_digital_employee tag field and the Digital Employee ID. When a user @DigitalEmployee in a group, the IM message will automatically carry this tag.

[0082] 2. Intelligent Service Trigger Mechanism: The IM server synchronizes messages marked with @_digital_employee to the intelligent service. After parsing the mark, the intelligent service extracts the message context (including group ID and historical conversation fragments) and triggers a dedicated Q&A intent stream.

[0083] 3. Question-and-answer process integrating multi-source knowledge:

[0084] Step 1: Filter out irrelevant content: Filter out meaningless content (such as "Received" or "Okay");

[0085] Step 2: Issue Validation: Remove invalid text other than issues, such as @XXXX;

[0086] Step 3: RAG Recall: Invoke the RAG service and customize the recall scope according to the supergroup scenario (configurable as "group-specific ES knowledge base + enterprise general knowledge base + FAQ").

[0087] Step 4: Results Summary: Input the multiple results retrieved by RAG into the large model to generate a concise and consistent response;

[0088] Step 5: Fallback: If RAG has no matching results, call the large model to generate a basic answer;

[0089] Step 6: Message Reply: The intelligent service calls the IM service to send a reply to the user as the @ed digital intelligence employee.

[0090] IV. Construct a dual-dimensional efficiency improvement mechanism of "real-time + scheduled" to solve the efficiency improvement problem of super group conversations.

[0091] Centered on the "Super Group Assistant" digital employee, efficiency within groups is improved through real-time analysis and scheduled tasks:

[0092] 1. Real-time business intent identification and triggering (combined with...) Figure 1 (As shown)

[0093] (1) Asynchronous message scheduling: After the IM server synchronizes the super group messages to the knowledge base service, the knowledge base service, in addition to storing data, transforms the messages into "intent analysis tasks" and pushes them to the Kafka queue (avoiding the performance pressure caused by a large number of messages directly calling the intelligent service in a short period of time).

[0094] (2) Real-time intent analysis: The intelligent service pulls tasks from the Kafka queue, calls "Real-time Intent Analysis Intent Stream", and uses the preset Prompt skill to call the large model to determine whether the message contains business intent (such as "create schedule" or "query expense reimbursement"). If the large model recognizes the business intent, it returns a structured result (example below):

[0095] {

[0096] "need_skill": true,

[0097] "skill_type": "schedule",

[0098] "parameters": {

[0099] "time": "2024-06-01 14:00",

[0100] "title": "Project Progress Meeting",

[0101] "location": "Online Meeting Room",

[0102] "participants": ["user1@company.com", "user2@company.com"]

[0103] }

[0104] }

[0105] (3) Business process triggering: The intelligent service assembles IM business message cards based on the results, calls the IM service to reply to the user as a super group assistant, and the user can trigger the corresponding business process by clicking the card:

[0106] 2. Scheduled automated services

[0107] (1) Scheduled task initialization: When a super group is created, the system automatically registers periodic scheduled tasks for the super group assistant (the group owner / administrator can switch tasks on and off and adjust the period through the group settings, such as 30 minutes / 1 hour). The task information is stored in the scheduled task scheduling center.

[0108] (2) Asynchronous task execution: After the scheduled task is triggered, the scheduling center pushes the task to the Kafka queue, and the intelligent service pulls the task and calls the "scheduled task intent stream";

[0109] (3) Contextualized message processing: (combined with) Figure 2 (As shown)

[0110] Step 1: Strategy Analysis: Based on the supergroup's preset skill strategy, determine the processing scenarios for scheduled tasks (such as "@message unresponsive reminder" or "dialogue summary").

[0111] Step 2: Message Grouping: Filter group messages within a specified time range according to scenario requirements—for example, in the "@Message Not Response Alert" scenario, select messages from "current time - 45 minutes" to "current time - 15 minutes" (leaving a 15-minute response window); in the "Conversation Summary" scenario, select messages from "current time - 30 minutes" to "current time";

[0112] Step 3: Scene matching and execution: For each message group, a dedicated Prompt is used to call the large model to determine whether it matches the scene (e.g., "whether there are any unresponsive @ messages"). If it matches, the corresponding operation is executed (e.g., sending a reminder or generating a summary), and the result is pushed through the IM service.

[0113] 5. Design a solution of "unified rendering + variable messages + two-way redirection" to solve the problem of seamless communication and business integration.

[0114] 1. Unified business message rendering across clients

[0115] (1) Standardized JSON template definition: Design a unified business message rendering template, define component types (type field, such as text / image / button / chart) and exclusive attributes (such as action_type of button, i18n_key of text) in JSON format, and add common control fields (i18n for multi-language adaptation, auth for permission control).

[0116] (2) Client-side parsing and rendering: After receiving the business message, each client (Android, iOS, HarmonyOS, H5) parses the JSON template and renders the UI according to its own technology stack (such as XML for Android and Storyboard for iOS) to ensure that the display style is consistent across different platforms.

[0117] 2. Business state-driven message mutable mechanism

[0118] (1) IM service interface extension: Add a "message modification interface" to the IM service to support updating message content (such as the status and field values ​​of business cards) by message ID;

[0119] (2) Status synchronization trigger: When the status of the business system changes (such as "schedule time modification" or "approval approved"), the business system calls the "message modification interface" of the IM service and passes in the updated business message JSON;

[0120] (3) Client status synchronization: When the client is online, it receives message updates in real time through the IM long connection; when offline, it retrieves unsynchronized message updates through the REST interface on the next login to ensure that business messages and business status are consistent in real time.

[0121] 3. Seamless integration of communication and business processes.

[0122] (1) Business redirection capability definition: In the button component of the business message template, define the action event type, which supports:

[0123] openUrl: Opens the business H5 page;

[0124] openApp: Redirects to the specified business page within the enterprise app;

[0125] openAgent: Redirects to the independent interactive page for digital employees;

[0126] requestApi: Directly calls the business interface to perform operations (such as "Approval Approved");

[0127] (2) Two-way connection: When a user clicks a button in a business message, the client parses the action type and triggers the corresponding jump or interface call to achieve direct connection between "communication scenario → business processing scenario"; at the same time, after the business system operation is completed, the status of the business message in the group can be updated through the "message variable mechanism" to achieve status synchronization between "business scenario → communication scenario".

[0128] The following is in conjunction with the appendix Figure 3 Described from the client's perspective:

[0129] This embodiment provides an agent-based instant messaging group collaboration method applied to a client, the method comprising:

[0130] Step S100: Send a supergroup creation request to the instant messaging server, triggering the instant messaging server to automatically bind a default digital employee as the supergroup assistant.

[0131] Specifically, as mentioned above, digital employees are marked with exclusive fields and are defined with exclusive UI interaction rules, address book permissions, and business capability scope; in addition to the bound default digital employees, other business-type digital employees are added as needed through the address book selection component within the enterprise's digital employee permission scope; the instant messaging server processes the sending and receiving of messages by digital employees according to user ID.

[0132] Step S200: When a conversation message is generated in the super group, the message is pushed to the instant messaging server, and the instant messaging server is triggered to send the message data to the intelligent server. The intelligent server is triggered to perform real-time intent analysis through the intent analysis process to determine whether the message contains business intent, and when the business intent is identified, it is assembled into business message data.

[0133] Preferably, in step S200, triggering the instant messaging server to send message data to the intelligent server includes: triggering the instant messaging server to synchronously push the message data to the search engine database, triggering the search engine database to store the message data, convert the message into an intent analysis task, and sending the intent analysis task to the intelligent server through synchronous push or asynchronous scheduling.

[0134] The search engine database is preferably Elasticsearch (ES) as described above. Furthermore, it is preferred to send data via asynchronous scheduling, with the asynchronous scheduling preferably using the Kafka queue method as described above.

[0135] For specific real-time message analysis reference Figure 1 As shown, and with reference to the real-time business intent recognition and triggering method described above.

[0136] In addition, the business message card includes a button component, which defines an action event type and supports:

[0137] openUrl: Opens the business H5 page;

[0138] openApp: Redirects to the specified business page within the enterprise app;

[0139] openAgent: Redirects to the independent interactive page for digital employees;

[0140] requestApi: Directly calls the business interface to perform operations;

[0141] When a user clicks a button in a business message card, the action type is parsed, and the corresponding redirect or API call is triggered.

[0142] When the business system status changes, the client receives the updated business message JSON received through the instant messaging server's message modification interface when the business system status changes, and displays the updated status on the business message card.

[0143] The business messages received by the client are processed using a unified cross-client business message rendering method, including: using a unified business message rendering template, defining component types and exclusive attributes in JSON format, and adding common control fields;

[0144] After receiving business message data, each client parses the JSON template and renders the UI according to its own technology stack to ensure consistent display style across different clients.

[0145] Step 300: Receive business message data from the intelligent server through the instant messaging server, reply as a super group assistant and display it in the form of a business message card. The business message card is set to trigger the corresponding business process by clicking on it.

[0146] The aforementioned agent-based instant messaging group collaboration method applied to clients may also include timed automated service steps;

[0147] The scheduled automation service includes:

[0148] When a super group creation request is sent to the instant messaging server, the instant messaging server is triggered to automatically register a periodic scheduled task for the super group assistant and store the task information in the scheduled task scheduling center.

[0149] After the scheduled task is triggered, it requests the scheduling center to send the task to the intelligent server via synchronous push or asynchronous scheduling.

[0150] And trigger the intelligent server to call the scheduled task intent stream for contextual message processing;

[0151] Receive contextualized message processing results from the intelligent server via the instant messaging server, and reply as a super group assistant.

[0152] Preferably, the scenario-based message processing includes: determining the processing scenario of the scheduled task according to the supergroup's preset skill strategy; filtering group messages within a specified time range according to scenario requirements; and for each message group, calling the large model through a dedicated Prompt to determine whether it matches the scenario, and if it matches, executing the corresponding operation.

[0153] The scenario requirements include handling user non-response to @ messages, summarizing periodic conversations, or handling tasks assigned within a group.

[0154] The aforementioned agent-based instant messaging group collaboration method applied to the client also includes a group-specific intelligent question-and-answer step, wherein the group-specific intelligent question-and-answer includes:

[0155] When a message @ Digital Employee is received in the super group, the message is synchronized to the intelligent server through the instant messaging server; and the intelligent server is requested to parse the message, trigger an intent stream based on the parsing result, and obtain the processing result by executing the intent stream;

[0156] Receive processing results from the intelligent server via the instant messaging server and receive a reply from the user who was tagged as a digital intelligence employee.

[0157] Specifically, the intelligent server parses the message, triggers an intent stream based on the parsing result, and obtains the processing result by executing the intent stream, including:

[0158] The intelligent server parses message tags, extracts message context, and triggers a dedicated question-and-answer intent stream. This dedicated intent stream includes: filtering meaningless content; removing invalid text other than the question; calling the RAG service to customize the recall scope according to the supergroup scenario, where the recall scope includes: the search engine database corresponding to the supergroup, the enterprise general knowledge base, and / or the FAQ database; inputting multiple results retrieved by RAG into a large model to generate a concise and unified answer; if RAG has no matching results, calling the large model to generate a basic answer.

[0159] In addition, this embodiment also provides a computer device, the device including: a processor and a memory; the memory is used to store one or more program instructions; the processor is used to run one or more program instructions to perform the steps of the agent-based instant messaging group collaboration method described above.

[0160] In addition, this embodiment also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the agent-based instant messaging group collaboration method described above.

[0161] Furthermore, this embodiment also provides a computer program product, which includes computer program instructions that, when executed by a processor, implement the steps of the agent-based instant messaging group collaboration method described above.

[0162] In this embodiment of the invention, the processor can be an integrated circuit chip with signal processing capabilities. The processor can be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in this embodiment of the invention. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in this embodiment of the invention can be directly implemented by a hardware decoding processor, or implemented by a combination of hardware and software modules in the decoding processor. The software modules can be located in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The processor reads information from the storage medium and, in conjunction with its hardware, completes the steps of the above methods. The storage medium can be memory, for example, volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. Non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDRSDRAM), Enhanced Synchronous DRAM (ESDRAM), Synchlink DRAM (SLDRAM), and Direct Rambus RAM (DRRAM).The storage media described in the embodiments of this invention are intended to include, but are not limited to, these and any other suitable types of memory. Those skilled in the art will recognize that the functions described in one or more of the above examples can be implemented using a combination of hardware and software. When applied software, the corresponding functions can be stored in a computer-readable medium or transmitted as one or more instructions or code on a computer-readable medium. Computer-readable media include computer storage media and communication media, wherein communication media include any medium that facilitates the transmission of computer programs from one place to another. Storage media can be any available medium accessible to general-purpose or special-purpose computers. Although the invention has been described in detail above with general description and specific embodiments, modifications or improvements can be made to it, which will be apparent to those skilled in the art. Therefore, all such modifications or improvements made without departing from the spirit of the invention are within the scope of protection claimed by this invention.

[0163] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Any simple modifications, equivalent changes, or alterations made by those skilled in the art using the disclosed technical content shall fall within the protection scope of the present invention.

Claims

1. A real-time communication group collaboration method based on intelligent agents, characterized in that, Applied to a client, the method includes: Send a supergroup creation request to the instant messaging server, triggering the instant messaging server to automatically bind a default digital employee as the supergroup assistant; When a conversation occurs in the super group, the message is pushed to the instant messaging server, which then sends the message data to the intelligent server. The intelligent server is then triggered to perform real-time intent analysis through the intent analysis process to determine whether the message contains business intent. When the business intent is identified, it is assembled into business message data. The system receives business message data from the intelligent server via an instant messaging server, replies as a super group assistant, and displays the message in the form of a business message card. The business message card is set to trigger the corresponding business process by clicking on it.

2. The agent-based instant messaging group collaboration method according to claim 1, characterized in that, The step of triggering the instant messaging server to send message data to the intelligent server includes: The instant messaging server is triggered to synchronously push message data to the search engine database, which in turn triggers the search engine database to store the message data, convert the message into an intent analysis task, and send the intent analysis task to the intelligent server via synchronous push or asynchronous scheduling.

3. The agent-based instant messaging group collaboration method according to claim 1, characterized in that, Digital employees are tagged with exclusive fields and have their own UI interaction rules, address book permissions, and business capability scope defined. In addition to the default digital employees that are bound to the system, other business-related digital employees can be added as needed through the contact list selection component, within the scope of the enterprise's digital employee permissions. The instant messaging server processes the sending and receiving of messages for digital employees based on user ID.

4. The agent-based instant messaging group collaboration method according to claim 1, characterized in that, The business message card includes a button component, which defines an action event type. Supports: openUrl: Opens a business H5 page; openApp: Redirects to a specified business page within the enterprise app; openAgent: Redirects to a separate interactive page for digital employees; requestApi: Directly calls business interfaces to perform operations; When a user clicks a button in a business message card, the action type is parsed, and the corresponding redirect or API call is triggered. Alternatively, when the business system status changes, the client receives the updated business message JSON received through the instant messaging server's message modification interface when the business system status changes, and displays the updated status on the business message card; Alternatively, the business messages received by the client are processed using a unified cross-client business message rendering method, including: using a unified business message rendering template, defining component types and exclusive attributes in JSON format, and adding common control fields; After receiving business message data, each client parses the JSON template and renders the UI according to its own technology stack to ensure consistent display style across different clients.

5. The agent-based instant messaging group collaboration method according to claim 1, characterized in that, It also includes a scheduled automation service, which includes: When a super group creation request is sent to the instant messaging server, the instant messaging server is triggered to automatically register a periodic scheduled task for the super group assistant and store the task information in the scheduled task scheduling center. After the scheduled task is triggered, it requests the scheduling center to send the task to the intelligent server via synchronous push or asynchronous scheduling. And trigger the intelligent server to call the scheduled task intent stream for contextual message processing; Receive contextualized message processing results from the intelligent server via the instant messaging server, and reply as a super group assistant.

6. The agent-based instant messaging group collaboration method according to claim 5, characterized in that, The scenario-based message processing includes: determining the processing scenario of the scheduled task based on the supergroup's preset skill strategy; filtering group messages within a specified time range according to scenario requirements; and for each message group, calling the large model through a dedicated Prompt to determine whether it matches the scenario, and executing the corresponding operation if it does. The scenario requirements include handling user non-response to @ messages, summarizing periodic conversations, or handling tasks assigned within a group.

7. The agent-based instant messaging group collaboration method according to claim 1, characterized in that, It also includes group-specific intelligent Q&A, which includes: When a message @a digital employee is received in the supergroup, the message is synchronized to the intelligent server via the instant messaging server. The intelligent server is then requested to parse the message, trigger an intent stream based on the parsing result, and obtain the processing result by executing the intent stream. The intelligent server's parsing of the message, triggering the intent stream based on the parsing result, and obtaining the processing result by executing the intent stream includes: parsing message tags, extracting message context, and triggering a dedicated question-and-answer intent stream; the dedicated question-and-answer intent stream includes: filtering meaningless content; removing invalid text other than questions; calling the RAG service to customize the recall scope according to the supergroup scenario, the recall scope including: the search engine database corresponding to the supergroup, the enterprise general knowledge base, and / or the FAQ database; inputting multiple results recalled by RAG into a large model to generate a concise and unified answer; if RAG has no matching results, calling the large model to generate a basic answer. Receive processing results from the intelligent server via the instant messaging server and receive a reply from the user who was tagged as a digital intelligence employee.

8. A computer device, characterized in that, The device includes: a processor and a memory; The memory is used to store one or more program instructions; The processor is configured to run one or more program instructions to perform the steps of the agent-based instant messaging group collaboration method as described in any one of claims 1-7.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the agent-based instant messaging group collaboration method as described in any one of claims 1-7.

10. A computer program product, characterized in that, The computer program product includes computer program instructions that, when executed by a processor, implement the steps of the agent-based instant messaging group collaboration method as described in any one of claims 1-7.

Images