Business problem processing method and device, equipment, medium and program product

Abstract

The invention provides a business problem processing method and device, equipment, a medium and a program product, can be applied to the field of artificial intelligence and the field of financial science and technology, and relates to application of a large model in an intelligent question and answer scene. The method comprises the following steps: acquiring a business event list corresponding to a business problem; wherein the business event list comprises event description information and business information; under the condition that the business information meets an automatic processing condition, calling a business knowledge base to process event description information through a model service component, and positioning a business scene and a reply template; wherein the model service component is a server tool constructed based on a model context protocol; calling a dynamic database through the model service component, and obtaining business data corresponding to the business scene; and based on the reply template, the business data and the event description information, calling the first large model through the model interface, and generating a question reply corresponding to the business event list.

Description

Technical Field

[0001] This application relates to the fields of artificial intelligence and fintech, and to the application of large models in intelligent question-answering scenarios. More specifically, it relates to a business problem processing method, apparatus, device, medium, and program product. Background Technology

[0002] Quick payment covers multiple payment scenarios and is characterized by multiple institutions and channels. The combined effect of the complexity of these scenarios and the diversity of participating entities leads to a continuous increase in business issues. The high number of business issues also results in their diversity and high complexity. Since quick payment is a key business for banks and has a large customer base, it is necessary to develop response forms for each business issue. The accuracy and timeliness of these responses directly impact customer satisfaction with the bank.

[0003] Currently, handling business issues still relies on manual processing of incident tickets. This involves multiple sources and applications, making it difficult to pinpoint business problems. Incident tickets are repeatedly transferred between business and technology departments, reducing the efficiency of responding to business incident tickets, leading to decreased customer satisfaction and increasing the risk of customer churn. Summary of the Invention

[0004] In view of the above problems, this application provides business problem processing methods, apparatus, equipment, media and program products to improve response efficiency.

[0005] According to a first aspect of this application, a business problem processing method is provided, comprising: obtaining a business event form corresponding to a business problem; wherein the business event form includes event description information and business information; when the business information meets automatic processing conditions, processing the event description information by calling a business knowledge base through a model service component to locate the business scenario and response template; wherein the model service component is a server-side tool built based on a model context protocol; obtaining business data corresponding to the business scenario by calling a dynamic database through the model service component; and generating a problem response corresponding to the business event form by calling a first main model through a model interface based on the response template, the business data, and the event description information.

[0006] According to an embodiment of this application, the model service component includes: a first model service component and a second model service component; wherein, the first model service component includes a first tool component, and the second model service component includes a second tool component.

[0007] According to an embodiment of this application, the step of calling the business knowledge base through the model service component to process the event description information, locate the business scenario and the response template includes: based on the first tool component, concatenating the event description information and preset supplementary description information; calling the knowledge base interface corresponding to the business knowledge base to send the concatenated description information to the business knowledge base; and based on the concatenated description information, matching it in the business knowledge base through a second major model to locate the business scenario and the response template.

[0008] According to an embodiment of this application, the step of calling the dynamic database through the model service component to obtain business data corresponding to the business scenario includes: obtaining the database query statement of the business scenario based on the second tool component; wherein the second tool component corresponds one-to-one with the business scenario; and connecting to the dynamic database based on the second tool component, and querying the database query statement according to the dynamic database to obtain the business data.

[0009] According to an embodiment of this application, the method further includes: acquiring historical business scenarios and historical response templates; generating a table file based on the structural information of the historical business scenarios and the historical response templates; converting the table file into text data and cutting the text data to obtain text fragments; converting the text fragments into fragment vectors and storing the fragment vectors in a vector database; and constructing the business knowledge base based on the vector database and embedding the second major model into the business knowledge base.

[0010] According to an embodiment of this application, the step of generating a question response corresponding to the business event order by calling a first major model through a model interface based on the response template, the business data, and the event description information includes: generating a constraint statement based on the response template, the business data, and the constraints; and processing the event description information by calling the first major model through the model interface based on the constraint statement to generate a question response corresponding to the business event order.

[0011] A second aspect of this application provides a business problem processing apparatus, comprising: an event ticket acquisition module, used to acquire a business event ticket corresponding to a business problem; wherein the business event ticket includes event description information and business information; a scenario and template positioning module, used to, when the business information meets the automatic processing conditions, call a business knowledge base through a model service component to process the event description information and locate the business scenario and response template; wherein the model service component is a server-side tool built based on a model context protocol; a business data acquisition module, used to call a dynamic database through the model service component to acquire business data corresponding to the business scenario; and a problem response generation module, used to, based on the response template, the business data, and the event description information, call a first main model through a model interface to generate a problem response corresponding to the business event ticket.

[0012] A third aspect of this application provides an electronic device comprising: one or more processors; and a memory for storing one or more computer programs, wherein the one or more processors execute the one or more computer programs to implement the steps of the method described above.

[0013] A fourth aspect of this application also provides a computer-readable storage medium having a computer program or instructions stored thereon, which, when executed by a processor, implement the steps of the above-described method.

[0014] The fifth aspect of this application also provides a computer program product, including a computer program or instructions that, when executed by a processor, implement the steps of the above-described method.

[0015] In the embodiments of this application, when the business information meets the conditions for automatic processing, the model service component built based on the model context protocol is used to call the business knowledge base to locate the scenario and response template, retrieve the business data from the dynamic database, and generate a response through the model interface in conjunction with the large model. This avoids the multi-department flow of manual processing, solves the problem of difficulty in locating problems caused by multiple sources and applications, improves the accuracy and efficiency of responses to quick payment business events, ensures the timeliness of responses, thereby improving customer satisfaction, reducing the risk of customer churn, and adapting to the complex and diverse business problem handling needs of quick payment. Attached Figure Description

[0016] The above-mentioned contents, other objects, features and advantages of this application will become clearer from the following description of embodiments with reference to the accompanying drawings, in which:

[0017] Figure 1 The illustrations depict application scenarios of business problem processing methods, apparatuses, devices, media, and program products according to embodiments of this application.

[0018] Figure 2 A flowchart illustrating a business problem processing method according to an embodiment of this application is shown schematically.

[0019] Figure 3 This illustration schematically shows an MCP component diagram of a business problem processing method according to an embodiment of this application;

[0020] Figure 4 This illustration schematically shows a quick business processing flowchart of a business problem handling method according to an embodiment of this application;

[0021] Figure 5 This schematically illustrates a structural block diagram of a business problem processing apparatus according to an embodiment of this application; and

[0022] Figure 6 A block diagram schematically illustrates an electronic device suitable for implementing a business problem processing method according to an embodiment of this application. Detailed Implementation

[0023] The embodiments of this application will now be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of this application. In the following detailed description, numerous specific details are set forth to provide a thorough understanding of the embodiments of this application for ease of explanation. However, it will be apparent that one or more embodiments may be implemented without these specific details. Furthermore, descriptions of well-known structures and technologies are omitted in the following description to avoid unnecessarily obscuring the concepts of this application.

[0024] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of this application. The terms “comprising,” “including,” etc., as used herein indicate the presence of features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0025] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0026] When using expressions such as "at least one of A, B and C", they should generally be interpreted in accordance with the meaning that is commonly understood by those skilled in the art (e.g., "a system having at least one of A, B and C" should include, but is not limited to, a system having A alone, a system having B alone, a system having C alone, a system having A and B, a system having A and C, a system having B and C, and / or a system having A, B and C, etc.).

[0027] With the rapid development of electronic payment in China, electronic payment has gradually begun to replace traditional paper money as a mainstream payment method. As a result, quick payment services on the banking side have become popular, with the number of customers and the volume of business increasing year by year, becoming one of the key businesses for banks. Consequently, business problems have also continued to grow.

[0028] Currently, the traditional business event ticket processing workflow based on the technology management workbench is as follows: Business departments create business event tickets on the technology management workbench based on customer complaints regarding quick payment scenarios. These tickets are then transferred to the technology department, which uses the information provided to determine the relevant quick payment institution and channel, identify the application to which the problem pertains, and forward it to the application side for processing. The application side uses the relevant information to determine the detailed business scenario of the quick payment transaction, locate the specific data source, query transaction messages, and determine the root cause of the problem, then develops a solution to respond to the business event ticket. This traditional workflow relies heavily on manual processing. Business departments may be dissatisfied with the response and return the ticket, leading to repetitive steps, increasing the burden on the technology department, and reducing the efficiency of event ticket responses.

[0029] Meanwhile, with the widespread use of quick payment transactions, the increase in the number of customers inevitably leads to a rise in customer complaints. The existing technologies still rely on manual processing of incident tickets, resulting in slow response times and low processing efficiency, which easily leads to more customer dissatisfaction. Secondly, due to the numerous detailed business scenarios under quick payment transactions, involving multiple applications, services, and data sources, it is difficult to pinpoint business problems. In addition, responding to business incident tickets requires cooperation between multiple departments, inevitably creating information gaps. The technology side's responses sometimes fail to meet the needs of the business side, ultimately causing incident tickets to repeatedly circulate between the business and technology departments.

[0030] This application provides a method for handling business issues, which involves obtaining a business event form corresponding to a business issue. The business event form includes event description information and business information. When the business information meets the conditions for automatic processing, a model service component calls a business knowledge base to process the event description information, locate the business scenario and response template. The model service component is a server-side tool built on a model context protocol. The model service component calls a dynamic database to obtain business data corresponding to the business scenario. Based on the response template, business data, and event description information, a first-level model is called through a model interface to generate a response to the business event form. In this application, when the business information meets the conditions for automatic processing, the model service component built on a model context protocol calls a business knowledge base to locate the scenario and response template, retrieves business data from a dynamic database, and generates a response by linking the first-level model through a model interface. This avoids the multi-department workflow of manual processing, solves the difficulty of problem location caused by multiple sources and applications, improves the accuracy and efficiency of responses to quick payment business event forms, ensures timely responses, thereby improving customer satisfaction, reducing customer churn risk, and adapting to the complex and diverse business issue handling needs of quick payment.

[0031] Figure 1 The illustration shows an application scenario diagram of a business problem processing method, apparatus, device, medium, and program product according to embodiments of this application.

[0032] like Figure 1 As shown, application scenario 100 according to this embodiment may include a first terminal device 101, a second terminal device 102, a third terminal device 103, a network 104, and a server 105. The network 104 serves as a medium for providing a communication link between the first terminal device 101, the second terminal device 102, the third terminal device 103, and the server 105. The network 104 may include various connection types, such as wired or wireless communication links, or fiber optic cables, etc.

[0033] Users can use the first terminal device 101, the second terminal device 102, and the third terminal device 103 to interact with the server 105 via the network 104 to receive or send messages, etc. Various communication client applications can be installed on the first terminal device 101, the second terminal device 102, and the third terminal device 103, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, social media platform software, etc. (for example only).

[0034] The first terminal device 101, the second terminal device 102, and the third terminal device 103 can be various electronic devices with displays and support web browsing, including but not limited to smartphones, tablets, laptops, and desktop computers.

[0035] Server 105 can be a server that provides various services, such as a backend management server that supports websites browsed by users using the first terminal device 101, the second terminal device 102, and the third terminal device 103 (this is just an example). The backend management server can analyze and process data such as received user requests, and feed back the processing results (such as web pages, information, or data obtained or generated according to user requests) to the terminal devices.

[0036] It should be noted that the business problem handling method provided in this application embodiment can generally be executed by server 105. Correspondingly, the business problem handling device provided in this application embodiment can generally be located in server 105. The business problem handling method provided in this application embodiment can also be executed by a server or server cluster that is different from server 105 and capable of communicating with the first terminal device 101, the second terminal device 102, the third terminal device 103, and / or server 105. Correspondingly, the business problem handling device provided in this application embodiment can also be located in a server or server cluster that is different from server 105 and capable of communicating with the first terminal device 101, the second terminal device 102, the third terminal device 103, and / or server 105.

[0037] It should be understood that Figure 1 The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included.

[0038] The following will be based on Figure 1 The described scene, through Figures 2-4 The business problem handling method according to the embodiments of this application will be described in detail.

[0039] Figure 2 A flowchart illustrating a business problem processing method according to an embodiment of this application is shown.

[0040] like Figure 2 As shown, the business problem handling method of this embodiment includes operations S210 to S240. This business problem handling method is not limited to a specific executing entity. The executing entity can be any electronic device, such as a terminal device or a server device, etc. The executing entity can also be any software application or client.

[0041] In operation S210, obtain the business event form corresponding to the business problem; the business event form includes event description information and business information.

[0042] The business side raises customer complaints about business issues, and the technology side processes and responds to these issues using a Large Language Model (LLM) to replace manual responses. The responses to business incidents are based on both real-time data sources (dynamic databases storing transaction details) and static data sources (business knowledge bases storing all financial transaction scenarios and response guidelines). Therefore, the LLM model's data sources possess both high real-time performance and multi-source capability. Traditional LLM models are trained on static data sources and cannot directly replace manual processing. Furthermore, to access multiple data sources, each data source requires independent interface development, making the solution design complex. This application adopts the Model Context Protocol (MCP), which provides a universal interface that allows the LLM model to access multiple data sources to meet the needs of business issue processing.

[0043] The platform integrates various technology support services on the bank side, covering service function modules in multiple fields. This platform provides banking business and technology departments with functions for creating, tracking, and responding to business event tickets, facilitating collaborative processing of business event tickets and tracking of business issues raised by customers among various departments within the bank.

[0044] A business event ticket is used to track business issues raised by customers. A business event ticket has the following elements: the person who raised the event ticket (the business personnel who created the event ticket), the processing department (which department is currently processing the event ticket; after the event ticket is raised, the technology management workbench will assign the event ticket to the technology department, and after the technology department completes its processing and response, the technology management workbench will reassign the event ticket back to the business department), business information (customer data, whether it is a quick service), and event description information (the problem raised by the customer, and a general description of the business scenario).

[0045] The business side submits customer complaints and related information through the bank's collaborative platform interface. Upon receiving the complaint, the system triggers a data verification module to perform integrity and format checks on core fields such as customer data and business scenario descriptions. After successful verification, metadata such as the complainant and the initial processing department (technology department) is automatically populated according to preset rules, generating a structured business event form with a unique identifier. This form is then synchronously written to the platform database, and the event form status is updated to "pending processing." Finally, the event form identifier and data access permissions are returned to the technology side, completing the acquisition process.

[0046] It should be noted that in the technical solution of this application, the customer information (including but not limited to user personal information, user image information, user device information, such as location information) and data (including but not limited to data used for analysis, stored data, and displayed data) involved are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of the relevant data all comply with relevant laws, regulations, and standards, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation entry points for users to choose to authorize or refuse.

[0047] In scenarios involving automated decision-making using personal information, the methods, devices, and systems provided in this application all offer users corresponding entry points for choosing to agree to or reject the automated decision-making results. If the user chooses to reject, the process proceeds to the expert decision-making stage. Here, "automated decision-making" refers to the activity of automatically analyzing and evaluating an individual's behavioral habits, interests, or economic, health, and credit status through computer programs, and then making a decision. Here, "expert decision-making" refers to the activity of making decisions by personnel who specialize in a particular field, possess specialized experience, knowledge, and skills, and have reached a certain level of professional expertise.

[0048] When operating S220, if the business information meets the conditions for automatic processing, the business knowledge base is called through the model service component to process the event description information, locate the business scenario and response template; the model service component is a server-side tool built on the model context protocol.

[0049] If any business information meets any of the automatic processing conditions, it will enter the automatic processing flow. The model service component will automatically generate the corresponding question and answer for the business event ticket. Automatic processing conditions include valid quick processing identifier, customer data compliance, high business frequency, and stable process. If the automatic processing conditions are not met, the business event ticket information will be transferred to the specific personnel in the technology department for manual processing.

[0050] When business information meets the conditions for automatic processing, the process is as follows: First, the event description information, customer basic data, and business type tags are integrated into a structured input and passed to the model service component built based on the model context protocol. This component parses the input information according to the format agreed upon in the protocol, calls the business knowledge base interface, and uses semantic retrieval and scenario matching algorithms to locate the corresponding business scenario category and associated response template from the business knowledge base. During automatic processing, the model context protocol ensures efficient data interaction between the component and the knowledge base, ensuring that the retrieval results accurately map the core needs of the event. At the same time, the component verifies the scenario matching degree and template applicability. After completing the location, it outputs a standardized response template, providing a compliant and business-scenario-appropriate basic framework for subsequent calls to the first main model to generate the final response, thus realizing automated scenario recognition and template matching.

[0051] The Model Context Protocol (MCP), as an open-source license, is a general-purpose interface designed for efficiently obtaining the context information needed by a model. It aims to integrate large language models with external data sources and tools, establishing a secure, bidirectional connection between the large model and the data source. MCP consists of MCP Host, MCP Client, and MCP Server.

[0052] MCP Host: The interaction point between the user and the Artificial Intelligence (AI) model, equivalent to an operating system or browser. AI models include large models, multimodal generative models, reinforcement learning models, graph neural network models, etc. Its core tasks are: receiving user instructions (such as natural language questions and code debugging requests) and passing them to the AI ​​model; creating and managing multiple client instances, coordinating task states and contexts; and controlling local resource access permissions (such as read / write restrictions on enterprise private databases).

[0053] MCP Client: A communication proxy between the host and the server, similar to a browser tab. Its core tasks are: establishing a 1:1 persistent connection with the server and transmitting requests / responses via communication protocols; executing tool calls (such as sending structured query commands) and managing resource access (such as reading local files); and implementing dynamic routing to automatically select the optimal server based on the AI ​​model's needs (e.g., prioritizing the local server to reduce latency).

[0054] MCP Server (Server Side): A provider of resources and tools, equivalent to an Application Programming Interface (API) gateway. The server side includes data sources, executable functions of service tools (such as database queries and file writing), prompt templates (predefined task instruction templates), responses to client requests, execution of tools, and return of structured results (such as structured query result sets).

[0055] According to an embodiment of this application, the business problem processing method further includes building a business knowledge base. The process of building the business knowledge base includes: obtaining historical business scenarios and historical response templates; generating a table file based on the structural information of the historical business scenarios and historical response templates; converting the table file into text data and cutting the text data to obtain text fragments; converting the text fragments into fragment vectors and storing the fragment vectors in a vector database; and building a business knowledge base based on the vector database and embedding a second major model into the business knowledge base.

[0056] Because it is necessary to locate the business scenario of the event ticket and generate the response result based on the response template, it is necessary to build a relevant AI business knowledge base. Taking the quick payment scenario as an example, the process of building the AI ​​business knowledge base is as follows:

[0057] (1) Information collection: Based on the business scenarios and responses of the quick payment business event tickets in recent years, summarize common business scenarios and response templates, and obtain historical business scenarios and historical response templates.

[0058] (2) Constructing the original knowledge base files: Based on the summarized structural information, generate relevant spreadsheet documents. The structural information includes the structure of the business scenario and the structure of the response template.

[0059] For example, the structure of the business scenario is as follows:

[0060] "Business Scenario Name": Quick Payment Agreement Signing and Termination Scenario;

[0061] "Business Scenario Description": Users have questions about the signing and termination scenarios of the quick payment agreement with the designated institution and need to know the signing and termination times of the relevant quick payment agreement.

[0062] The structure of the response template is as follows:

[0063] "Business Scenario Name": Quick Payment Agreement Signing and Termination Scenario;

[0064] "Response Template": Upon inquiry, the quick payment agreement signed between the customer and {Customer Placeholder 1} was signed at {Time Placeholder 2} and terminated at {Time Placeholder 3}.

[0065] The response template contains fields to be filled in, which are placeholders. Based on the above business scenario and response template structure, generate the corresponding spreadsheet file.

[0066] (3) Document Conversion and Segmentation: Convert the table file into a plain text document and perform text segmentation to divide the document into fragments. First, use a document conversion tool to extract the core content such as the business scenario name, description, and response template from the table, and remove format redundancy to generate a plain text document. Then, perform text segmentation according to the logic of single scenario + corresponding template to ensure that each fragment focuses on a set of related information, avoid confusion between scenario and template, and obtain a clear, independent and complete text fragment.

[0067] (4) Document Vectorization and Storage: The segmented text fragments are converted into vectors and stored in a vector database. A pre-trained language model can be used to encode the segmented text fragments. The model captures the semantic associations and keyword features of the business scenario description and response template, mapping each text fragment to a dense vector of fixed dimensions (fragment vector). After the vectors are generated, they are standardized to optimize retrieval efficiency, and then imported into the vector database in batches. The database establishes an efficient index to store the mapping relationship between fragment vectors and corresponding text fragments, enabling fast vector querying and matching, and providing underlying data support for subsequent business scenario positioning and response generation.

[0068] (5) Constructing a business knowledge base: Using a vector database as the core carrier, a knowledge base architecture of scene retrieval, template matching, and result output is built. The vector data is associated with the original text fragments, and semantic similarity retrieval rules are configured to support the rapid location of matching historical scenes and response templates based on business event orders. At the same time, placeholder filling logic and update mechanisms are embedded to ensure that the knowledge base can be dynamically iterated, ultimately forming a business knowledge base that can efficiently support AI response generation.

[0069] (6) Embedding the Large Model: First, build the interaction interface between the large model and the business knowledge base, configure the semantic retrieval calling logic, so that the large model can trigger vector matching and scene positioning of the knowledge base based on business events. Then, embed the placeholder filling rules of the reply template and the business logic into the reasoning process of the large model, so that the large model can call the matching results to generate accurate replies. Finally, fine-tune and optimize the interaction efficiency through business data to achieve deep integration between the large model and the knowledge base, forming an end-to-end intelligent reply capability. Embedding the large model in the business knowledge base facilitates system calls, and enables the large model to return relevant information based on the document content after inputting a description.

[0070] In the embodiments of this application, a business knowledge base is constructed by standardizing historical business scenarios and historical response templates, and a large model is embedded to make the knowledge base data structured and vectorized, thereby improving the efficiency and accuracy of retrieval and matching, facilitating the rapid location of business scenarios and templates. At the same time, historical data optimizes the iteration of the knowledge base, further ensuring the accuracy of responses and the timeliness of event processing.

[0071] When operating S230, the model service component calls the dynamic database to obtain business data corresponding to the business scenario.

[0072] The model service component, based on the identified business scenario tags and in accordance with the Model Context Protocol (MCP) specification, generates a structured query request that explicitly includes a scenario identifier, data query fields (such as customer order details and business rule parameters), and data filtering conditions (key information such as customer identifiers and order numbers from associated business event forms). Subsequently, the component initiates a request through a pre-defined database interface. Upon receiving the request, the dynamic database performs an efficient retrieval based on the query conditions, matching the scenario-related data tables and extracting the target business data (including basic configuration data and historical interaction data). The MCP protocol ensures the accuracy of the query command and the security of data transmission, and the database synchronously returns structured data results. The component performs integrity verification and format standardization on the data to ensure a high degree of data compatibility with the business scenario. Finally, the standardized business data is output, providing data support for subsequent response template filling and model inference to generate accurate responses.

[0073] According to an embodiment of this application, the model service component includes: a first model service component and a second model service component; wherein, the first model service component includes a first tool component, and the second model service component includes a second tool component.

[0074] Figure 3 The diagram illustrates an MCP component diagram of a business problem handling method according to an embodiment of this application.

[0075] like Figure 3 As shown, a business event ticket is generated based on business issues through the technology management platform. This ticket is then transferred to the technology department, which connects to an automated system for handling quick payment business issues. This system consists of an MCP component, and the business MCP component connects to the technology management platform. Building the MCP component includes setting up MCP Host, MCP Client, and MCP Server components. The MCP Server also represents the model service component in this application. Data sources include all quick payment business scenarios, a full business database involving quick payment applications (stores relevant business information for quick transactions), and basic response templates for quick payment issues.

[0076] MCP HOST: Select an LLM model with an open interface (the primary model), which can generate text content via API calls. In actual use, the system will transmit the description to the LLM model through the API interface, enabling the LLM model to generate the specified content.

[0077] MCP Client: The core component of the system, used to implement business logic and call MCP HOST and MCPSERVER. Required input information is the event form description field INPUT. The processing flow is as follows: Based on the event form description field INPUT and the event description information, the T1 component of MCP SERVER1 is called to obtain the business scenario D0 and response template D1 in JSON format. Based on the obtained business scenario D0, the Tn component in MCP SERVER2 is called to obtain the relevant business data D2 in JSON format required by the business scenario. D1 and D2 are provided to the large model with added constraints, and the LLM model interface is called to generate the response.

[0078] MCP SERVER: This system requires the construction of two MCP model service components. MCP SERVER1 (the first model service component) is used to connect to the AI ​​business knowledge base to obtain business scenarios and response templates; MCP SERVER2 (the second model service component) is used to connect to the business database to provide real-time business data, which is used to generate the final response result by combining the template information.

[0079] The MCP SERVER setup consists of four components: Tool, Resource, Prompts, and Error Handler. The Tool component is the core component and at least one is required. The other components can be built according to needs. The Resource component is used for static resource acquisition and is not involved in this system. The Prompts component manages dialogue templates and prompts, but its templates are generally static. This system uses an AI business knowledge base for template acquisition, so it is also not involved. The Error Handler component provides unified error management and response. A simple setup is needed to acquire all system errors and return a unified response when a system error occurs: "System error, please transfer this event to manual processing. If you have any questions, contact the administrator."

[0080] In the embodiments of this application, the model service component is split into a first component and a second component containing different tool components, which can be precisely divided according to business scenarios and processing requirements. This component-based splitting achieves functional decoupling, improves the flexibility and adaptability of invocation, simplifies subsequent maintenance and iteration, and helps to efficiently locate business scenarios and retrieve data.

[0081] When operating S240, based on the response template, business data, and event description information, the first major model is called through the model interface to generate the question response corresponding to the business event.

[0082] The primary model is the LLM model configured in the MCP host. The LLM model can be selected from those based on a general natural language processing (NLP) model foundation, fine-tuned with data from specific vertical domains. By utilizing data from scenarios such as financial transactions and quick payments, the model can accurately grasp industry business logic and generate responses that comply with industry compliance requirements and process standards.

[0083] When calling the first major model, the necessary data, such as the response template of the business event ticket, business data, and event description information, are first encapsulated through a standardized interface, and the request format is adapted through the MCP protocol. Then, through the permission verification and resource scheduling module, the request is routed to the selected major model instance. After the major model inference is completed, the result is encapsulated and returned in a preset format, and the stability of data transmission and interaction compatibility are ensured through the interface protocol.

[0084] In the embodiments of this application, when the business information meets the conditions for automatic processing, the model service component built based on the model context protocol is used to call the business knowledge base to locate the scenario and response template, retrieve the business data from the dynamic database, and generate a response through the model interface in conjunction with the large model. This avoids the multi-department flow of manual processing, solves the problem of difficulty in locating problems caused by multiple sources and applications, improves the accuracy and efficiency of responses to quick payment business events, ensures the timeliness of responses, thereby improving customer satisfaction, reducing the risk of customer churn, and adapting to the complex and diverse business problem handling needs of quick payment.

[0085] According to an embodiment of this application, in operation S220, the event description information is processed by calling the business knowledge base through the model service component to locate the business scenario and response template, including: based on the first tool component, splicing the event description information and the preset supplementary description information; calling the knowledge base interface corresponding to the business knowledge base to send the spliced ​​description information to the business knowledge base; and based on the spliced ​​description information, matching in the business knowledge base through the second major model to locate the business scenario and response template.

[0086] MCP SERVER1 includes a first tool component (T1) used to query the internal business knowledge base to obtain the business scenarios and response templates corresponding to relevant descriptions. The input of T1 is the descriptive information Q used to match the business scenarios, such as event description information and supplementary description information. Q is provided to the AI ​​business knowledge base to match the appropriate business scenarios, and this input item is a required field.

[0087] The corresponding method F1 for component T1 is as follows:

[0088] Input: Field Q.

[0089] Method implementation: Field Q (event description information) + fixed question description supplementary field (supplementary description information), such as "Event description information xxxxxx, the above is a business scenario description, please provide the most suitable business scenario and response template for this description, and return it in JSON format". The two types of information are concatenated into a question and answer statement sent to the business knowledge base. The business knowledge base interface is called to use the second major model embedded in it to match the business scenario and response template in the knowledge base.

[0090] It is worth noting that the second major model is a natural language processing model embedded in the business knowledge base. This model can be fine-tuned based on the enterprise's private business data and serves only the internal business knowledge base. It can accurately match the internal business scenarios and response templates, while ensuring data security and adapting to the knowledge base embedding needs of sensitive fields such as finance and government affairs.

[0091] In the embodiments of this application, the event description and supplementary information are spliced ​​together by the first tool component, which enriches the dimensions of the problem description. Then, the knowledge base interface is called and the business scenario and response template are located by matching with the large model, which improves the accuracy of knowledge base retrieval. By combining the large model with the knowledge base, the business scenario involved in the problem can be located more quickly and accurately and the relevant data can be retrieved, which facilitates the response to complex business problems.

[0092] For example, taking the automated processing of quick business issues as an example, Figure 4 The diagram illustrates a quick business processing flowchart of a business problem handling method according to an embodiment of this application.

[0093] like Figure 4 As shown, the quick service processing flow includes steps 1 through 8, as detailed below:

[0094] Step 1: Customer complaints are handled by creating a business event form in the technology management workbench. The event form description and whether it is a quick service field are filled in as required.

[0095] Step 2: After the business department completes the business event order, the technology management platform will assign the event order to the technology department for processing.

[0096] Step 3: The business issue processing system will call the technology management work platform interface in real time to obtain the event tickets currently being processed by the technology department, along with the event description and business information. The condition for quick processing is that the business event ticket is a quick service; this can be determined by the quick service field. If it is not a quick service, no processing will be performed, and it will await manual processing. If it is a quick service, the event description information will be entered into the MCP Client for further processing.

[0097] Step 4: Based on the event description information in the event ticket description field, the MCP Client calls the T1 component of MCP SERVER1, uses the T1 component to call the AI ​​knowledge base, locates the relevant event ticket, and obtains the business scenario and response template in JSON format.

[0098] Step 5: After the business scenario and response template are determined, make a judgment based on the obtained business scenario, call the Tn component in MCPSERVER2, and use the Tn component to call the dynamic database (the data source related to the business scenario) to query the business data in JSON format required by the relevant business scenario, such as transaction details, transaction time, etc.

[0099] Step 6: Provide the response template and business data to the large model (such as a natural language model) and add constraints. Call the model interface, use the large model to perform data reasoning, and generate the final question response.

[0100] Step 7: Return the model's output answers to the technology management platform, which will then record the answers and return them to the business department.

[0101] Step 8: If the business department supplements the event ticket description or adds questions raised by the customer and forwards it to the technology department, then start processing the event ticket again from step 3, repeating steps 3 to 6 until the business department closes the event ticket.

[0102] The processing time of quick business event tickets has been optimized by converting manually processed event tickets to LLM model processing. This has improved the real-time response to business event tickets, optimized the processing time of quick business event tickets, effectively reduced customer complaint processing time, and improved customer satisfaction.

[0103] According to an embodiment of this application, in operation S230, the model service component calls the dynamic database to obtain business data corresponding to the business scenario, including: obtaining the database query statement of the business scenario based on the second tool component; wherein, the second tool component corresponds one-to-one with the business scenario; and connecting to the dynamic database based on the second tool component, and querying the database query statement according to the dynamic database to obtain business data.

[0104] The primary task of MCP SERVER2 is to acquire business data. Due to varying business scenarios, the databases to be queried and the related business statements differ. Therefore, different second tool components (Tn) need to be built for different business scenarios, although the overall structure of the tool components remains similar. Each business scenario requires a corresponding tool component to acquire the business data specific to that scenario.

[0105] The input for Tn consists of the fields necessary for querying the database; this input is a required field.

[0106] The corresponding method Fn for component Tn is as follows:

[0107] Input: The fields required to query the database.

[0108] Method implementation: Implement the specific SQL (Structured Query Language) statements and connect to the database to obtain relevant business information.

[0109] Output: Fields in JSON format, where the attribute names are consistent with the field names to be filled in the corresponding response template for the relevant business scenario.

[0110] In the embodiments of this application, based on the correspondence between tool components and business scenarios, the corresponding database query statements can be accurately obtained, and then the query can be executed by directly connecting to the dynamic database through the component. This avoids query statement adaptation deviation, improves the accuracy and efficiency of data retrieval, ensures timely acquisition of business data, provides reliable support for generating responses for large models, and reduces the cost of manual querying and verification.

[0111] According to an embodiment of this application, in operation S240, based on the response template, business data, and event description information, the first major model is called through the model interface to generate a question response corresponding to the business event, which includes: generating a constraint statement based on the response template, business data, and constraints; and based on the constraint statement, the first major model is called through the model interface to process the event description information and generate a question response corresponding to the business event.

[0112] The system parses the structure and placeholders of the response template, associates them with corresponding fields in the business data (such as the agreement institution, time, etc.), and generates constraint statements by combining compliance standards, expression style, and other constraints. Then, it concatenates the constraint statements, event description information, and key information of the template into standardized input, and calls the first major model in MCP HOST through the interface. The model integrates the input semantic data and template, fills in the placeholders according to the constraint logic, and generates accurate, compliant, and business-scenario-appropriate single response for the event.

[0113] For example, the constraints are: (1) Generate a response based on the response template and business data. (2) Optimize the response template appropriately to ensure fluency and readability, maintain a friendly tone of communication, and maintain professionalism and accuracy.

[0114] In the embodiments of this application, a large model is used to replace the manual analysis, processing, and response to quick business event tickets. Constraint statements are used to regulate the output boundaries of the large model to prevent responses from deviating from business rules. At the same time, core business elements are integrated to ensure that responses are both relevant to the actual scenario and accurate and compliant. This further improves the accuracy of event ticket responses, reduces manual verification costs, reduces the workload of operation and maintenance personnel, and ensures processing timeliness and customer satisfaction.

[0115] Based on the above-described business problem handling method, this application also provides a business problem handling apparatus. The following will be combined with... Figure 5 The device is described in detail.

[0116] Figure 5 A schematic block diagram of a business problem processing apparatus according to an embodiment of this application is shown.

[0117] like Figure 5 As shown, the business problem processing device 500 in this embodiment includes an event order acquisition module 510, a scene and template positioning module 520, a business data acquisition module 530, and a problem response generation module 540.

[0118] The event ticket acquisition module 510 is used to acquire the business event ticket corresponding to the business problem; wherein, the business event ticket includes event description information and business information. In one embodiment, the event ticket acquisition module 510 can be used to perform the operation S210 described above, which will not be repeated here.

[0119] The scenario and template location module 520 is used to locate the business scenario and response template by calling the business knowledge base through the model service component to process the event description information when the business information meets the automatic processing conditions; wherein, the model service component is a server-side tool built based on the model context protocol. In one embodiment, the scenario and template location module 520 can be used to perform the operation S220 described above, which will not be repeated here.

[0120] The business data acquisition module 530 is used to call the dynamic database through the model service component to obtain business data corresponding to the business scenario. In one embodiment, the business data acquisition module 530 can be used to perform the operation S230 described above, which will not be repeated here.

[0121] The question-and-answer generation module 540 is used to generate a question-and-answer corresponding to the business event based on the answer template, the business data, and the event description information, by calling the first main model through the model interface. In one embodiment, the question-and-answer generation module 540 can be used to perform the operation S240 described above, which will not be repeated here.

[0122] According to an embodiment of this application, the model service component includes: a first model service component and a second model service component; wherein, the first model service component includes a first tool component, and the second model service component includes a second tool component.

[0123] According to an embodiment of this application, the scenario and template positioning module 520 includes: a splicing unit, used to splice the event description information and preset supplementary description information based on the first tool component; a calling unit, used to call the knowledge base interface corresponding to the business knowledge base and send the spliced ​​description information to the business knowledge base; and a matching unit, used to match the spliced ​​description information in the business knowledge base through a second model to locate the business scenario and the response template.

[0124] According to an embodiment of this application, the business data acquisition module 530 includes: a statement acquisition unit, configured to acquire a database query statement for the business scenario based on the second tool component; wherein the second tool component corresponds one-to-one with the business scenario; and a query unit, configured to connect to the dynamic database based on the second tool component, and query the database query statement according to the dynamic database to acquire the business data.

[0125] According to an embodiment of this application, the apparatus 500 further includes: a knowledge base construction module, configured to acquire historical business scenarios and historical response templates; generate a table file based on the structural information of the historical business scenarios and the historical response templates; convert the table file into text data and cut the text data to obtain text fragments; convert the text fragments into fragment vectors and store the fragment vectors in a vector database; and construct the business knowledge base based on the vector database and embed the second major model into the business knowledge base.

[0126] According to an embodiment of this application, the question and answer generation module 540 includes: a constraint generation unit, used to generate constraint statements based on the answer template, the business data, and constraint conditions; and an answer generation unit, used to generate a question and answer corresponding to the business event by calling the first large model to process the event description information through the model interface based on the constraint statements.

[0127] According to embodiments of this application, any multiple modules among the event order acquisition module 510, scene and template positioning module 520, business data acquisition module 530, question answer generation module 540, and knowledge base construction module can be merged into one module, or any one of these modules can be split into multiple modules. Alternatively, at least some of the functions of one or more of these modules can be combined with at least some of the functions of other modules and implemented in one module. According to embodiments of this application, at least one of the event order acquisition module 510, scene and template positioning module 520, business data acquisition module 530, question answer generation module 540, and knowledge base construction module can be at least partially implemented as hardware circuits, such as field-programmable gate arrays (FPGAs), programmable logic arrays (PLAs), systems-on-a-chip, systems-on-a-substrate, systems-on-package, application-specific integrated circuits (ASICs), or any other reasonable means of integrating or packaging circuits, or implemented in hardware or firmware, or in any one of software, hardware, and firmware implementations, or in a suitable combination of any of these. Alternatively, at least one of the event order acquisition module 510, scenario and template positioning module 520, business data acquisition module 530, question answer generation module 540, and knowledge base construction module can be implemented at least partially as a computer program module, which can perform corresponding functions when the computer program module is run.

[0128] Figure 6 A block diagram schematically illustrates an electronic device suitable for implementing a business problem processing method according to an embodiment of this application.

[0129] like Figure 6 As shown, an electronic device 900 according to an embodiment of this application includes a processor 901, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 902 or a program loaded from a storage portion 908 into a random access memory (RAM) 903. The processor 901 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or an associated chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 901 may also include onboard memory for caching purposes. The processor 901 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of this application.

[0130] RAM 903 stores various programs and data required for the operation of electronic device 900. Processor 901, ROM 902, and RAM 903 are interconnected via bus 904. Processor 901 executes various operations of the method flow according to embodiments of this application by executing programs in ROM 902 and / or RAM 903. It should be noted that the programs may also be stored in one or more memories other than ROM 902 and RAM 903. Processor 901 may also execute various operations of the method flow according to embodiments of this application by executing programs stored in said one or more memories.

[0131] According to embodiments of this application, the electronic device 900 may further include an input / output (I / O) interface 905, which is also connected to a bus 904. The electronic device 900 may also include one or more of the following components connected to the input / output (I / O) interface 905: an input section 906 including a keyboard, mouse, etc.; an output section 907 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 908 including a hard disk, etc.; and a communication section 909 including a network interface card such as a LAN card, modem, etc. The communication section 909 performs communication processing via a network such as the Internet. A drive 910 is also connected to the input / output (I / O) interface 905 as needed. A removable medium 911, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 910 as needed so that computer programs read from it can be installed into the storage section 908 as needed.

[0132] This application also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or it may exist independently and not assembled into the device / apparatus / system. The computer-readable storage medium carries one or more programs, which, when executed, implement the method according to the embodiments of this application.

[0133] According to embodiments of this application, the computer-readable storage medium can be a non-volatile computer-readable storage medium, such as including but not limited to: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this application, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, according to embodiments of this application, the computer-readable storage medium may include ROM 902 and / or RAM 903 and / or one or more memories other than ROM 902 and RAM 903 described above.

[0134] Embodiments of this application also include a computer program product comprising a computer program containing program code for performing the methods shown in the flowchart. When the computer program product is run on a computer system, the program code enables the computer system to implement the business problem-solving methods provided in the embodiments of this application.

[0135] When the computer program is executed by the processor 901, it performs the functions defined in the system / apparatus of this application embodiment. According to the embodiments of this application, the systems, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0136] In one embodiment, the computer program may rely on a tangible storage medium such as an optical storage device or a magnetic storage device. In another embodiment, the computer program may also be transmitted and distributed in the form of signals over a network medium, and downloaded and installed via the communication section 909, and / or installed from a removable medium 911. The program code contained in the computer program can be transmitted using any suitable network medium, including but not limited to: wireless, wired, etc., or any suitable combination thereof.

[0137] In such an embodiment, the computer program can be downloaded and installed from a network via the communication section 909, and / or installed from the removable medium 911. When the computer program is executed by the processor 901, it performs the functions defined in the system of this application embodiment. According to the embodiments of this application, the systems, devices, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0138] According to embodiments of this application, program code for executing the computer programs provided in the embodiments of this application can be written in any combination of one or more programming languages. Specifically, these computational programs can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, languages ​​such as Java, C++, Python, "C", or similar programming languages. The program code can be executed entirely on the user's computing device, partially on the user's device, partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0139] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0140] Those skilled in the art will understand that the features described in the various embodiments of this application can be combined and / or combined in various ways, even if such combinations or combinations are not explicitly described in this application. In particular, the features described in the various embodiments of this application can be combined and / or combined in various ways without departing from the spirit and teachings of this application. All such combinations and / or combinations fall within the scope of this application.

Claims

1. A method for handling business problems, characterized in that, The method includes: Obtain the business event form corresponding to the business issue; wherein, the business event form includes event description information and business information; When the business information meets the conditions for automatic processing, the event description information is processed by calling the business knowledge base through the model service component to locate the business scenario and response template; wherein, the model service component is a server-side tool built based on the model context protocol; The model service component calls the dynamic database to obtain business data corresponding to the business scenario; and Based on the response template, the business data, and the event description information, the first major model is called through the model interface to generate the question response corresponding to the business event.

2. The method according to claim 1, characterized in that, The model service components include: a first model service component and a second model service component; wherein, the first model service component includes a first tool component, and the second model service component includes a second tool component.

3. The method according to claim 2, characterized in that, The step of calling the business knowledge base through the model service component to process the event description information, locate the business scenario and response template includes: Based on the first tool component, the event description information and preset supplementary description information are concatenated; Call the knowledge base interface corresponding to the business knowledge base and send the concatenated description information to the business knowledge base; and Based on the concatenated description information, the second major model is used to match the business knowledge base to locate the business scenario and the response template.

4. The method according to claim 2, characterized in that, The step of calling the dynamic database through the model service component to obtain the business data corresponding to the business scenario includes: The database query statement for the business scenario is obtained based on the second tool component; wherein, the second tool component corresponds one-to-one with the business scenario; and The second tool component connects to the dynamic database, and queries the database based on the dynamic database to obtain the business data.

5. The method according to any one of claims 3, characterized in that, The method further includes: Retrieve historical business scenarios and historical response templates; Based on the structure information of the historical business scenarios and the historical response templates, a table file is generated; The table file is converted into text data, and the text data is then segmented to obtain text fragments; The text fragments are converted into fragment vectors, and the fragment vectors are stored in a vector database; and The business knowledge base is constructed based on the vector database, and the second major model is embedded in the business knowledge base.

6. The method according to claim 1, characterized in that, The step of generating a question response corresponding to the business event order by calling the first major model through the model interface based on the response template, the business data, and the event description information includes: Generate constraint statements based on the response template, the business data, and the constraints; and Based on the constraint statement, the first large model is called through the model interface to process the event description information and generate the question answer corresponding to the business event order.

7. A business problem processing device, characterized in that, The device includes: The event ticket acquisition module is used to acquire business event tickets corresponding to business issues; wherein, the business event ticket includes event description information and business information; The scenario and template localization module is used to locate the business scenario and response template by calling the business knowledge base through the model service component to process the event description information when the business information meets the automatic processing conditions; wherein, the model service component is a server-side tool built based on the model context protocol; The business data acquisition module is used to call the dynamic database through the model service component to obtain business data corresponding to the business scenario; and The question and answer generation module is used to generate the question and answer corresponding to the business event based on the answer template, the business data and the event description information, by calling the first major model through the model interface.

8. An electronic device, comprising: One or more processors; Memory, used to store one or more computer programs. The characteristic feature is that the one or more processors execute the one or more computer programs to implement the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program or instructions stored thereon, characterized in that, When the computer program or instructions are executed by a processor, they implement the steps of the method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program or instructions, characterized in that, When the computer program or instructions are executed by a processor, they implement the steps of the method according to any one of claims 1 to 6.

Images