Business process operation and maintenance method and device, electronic equipment and computer program product

Abstract

The application discloses a kind of business process operation and maintenance method, device, electronic equipment and computer program product.It relates to artificial intelligence field, the method includes: in response to the maintenance instruction of target business process input by user, text prompt word is generated based on the original content of maintenance instruction;Text prompt word is semantically analyzed, and analysis result is obtained, and text prompt word is decomposed into multiple maintenance operation information based on analysis result;Corresponding maintenance task prompt word is generated based on each maintenance operation information, large language model is called to process maintenance task prompt word, and the policy maintenance instruction corresponding to each maintenance operation information is generated;Each policy maintenance instruction indicated service interface is called to execute multiple policy maintenance operations.By the present application, the problem of low efficiency of business process operation and maintenance in related art is solved.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence, and more specifically, to a method, apparatus, electronic device, and computer program product for the operation and maintenance of a business process. Background Technology

[0002] In the business process operation and maintenance platform of financial institutions, the policy management system in related technologies adopts a single-key direct connection architecture. The system architecture includes a request forwarding layer, a key authentication layer, and an atomic service layer. Its workflow is as follows: receiving a request, attaching a fixed key for authentication, synchronously waiting for a response, and returning the result. The technical implementation relies on direct client invocation of the basic Hypertext Transfer Protocol.

[0003] However, business personnel must manually configure strategy parameters one by one according to predetermined steps. This fully manual operation is inefficient and prone to errors due to human negligence. The system only provides static effect statistics, relying on manual data analysis before strategy adjustments are made. This linear approach results in poor response timeliness, making it impossible to dynamically optimize strategies based on real-time effects, leading to low operational efficiency in business processes.

[0004] There is currently no effective solution to the problem of low operational efficiency in business processes related to these technologies. Summary of the Invention

[0005] The main purpose of this application is to provide a method, apparatus, electronic device and computer program product for the operation and maintenance of business processes, so as to solve the problem of low efficiency in the operation and maintenance of business processes in related technologies.

[0006] To achieve the above objectives, according to one aspect of this application, a method for operating and maintaining a business process is provided. The method includes: responding to a maintenance instruction for a target business process input by a user; generating text prompts based on the original content of the maintenance instruction; performing semantic parsing on the text prompts to obtain parsing results; decomposing the text prompts into multiple maintenance operation information based on the parsing results; generating corresponding maintenance task prompts based on each maintenance operation information; calling a large language model to process the maintenance task prompts to generate a policy maintenance instruction corresponding to each maintenance operation information; and calling the service interface indicated by each policy maintenance instruction to execute multiple policy maintenance operations.

[0007] Optionally, generating text prompts based on the original content of maintenance instructions includes: converting images or speech in the original content into text content to obtain target text content; obtaining the user's historical conversation content, merging the historical conversation content with preset policy maintenance prompts to generate initial context; and constructing text prompts based on the initial context and target text content.

[0008] Optionally, semantic parsing of the text prompts yields the following results: processing the text prompts using a large language model to obtain strategy maintenance types, wherein the strategy maintenance types include at least one of the following: addition, deletion, and modification; extracting business parameters from the text prompts, wherein the business parameters include at least one of the following: target customer group, financial product, and time range; and determining the business parameters and strategy maintenance types as the parsing results.

[0009] Optionally, the text prompt words are decomposed into multiple maintenance operation information based on the parsing results, including: determining multiple service interfaces that need to be processed based on the policy maintenance type; determining the operation type for each service interface based on business parameters; and determining each service interface and the operation type corresponding to the service interface as a maintenance operation information, thereby obtaining multiple maintenance operation information.

[0010] Optionally, after executing multiple policy maintenance operations by calling the service interface indicated by each policy maintenance instruction, the method further includes: obtaining the policy execution indicator value of the target business process within the target period, and determining whether the policy execution indicator value is greater than or equal to the indicator value threshold; if the policy execution indicator value is greater than or equal to the indicator value threshold, issuing a prompt message, wherein the prompt message is used to indicate that the maintenance of the target business process is complete; if the policy execution indicator value is less than the indicator value threshold, optimizing the target business process in the policy knowledge base, wherein the policy knowledge base contains multiple business processes.

[0011] Optionally, optimizing the target business process in the strategy knowledge base includes: retrieving historical reference strategies from the strategy knowledge base, wherein the similarity between the business scenario of the historical reference strategies and the business scenario of the target business process is greater than or equal to a preset similarity threshold; inputting the historical reference strategies and strategy execution indicator values ​​into a large language model, processing them to obtain the strategy parameters to be adjusted and the recommended parameter values; setting the strategy parameters to be adjusted for the target business process to the recommended parameter values, thereby obtaining the optimized target business process.

[0012] Optionally, after decomposing the text prompt words into multiple maintenance operation information based on the parsing results, the method further includes: determining the execution order among the multiple policy maintenance operations; calling the service interface indicated by each policy maintenance instruction to execute the multiple policy maintenance operations includes: executing the multiple policy maintenance operations sequentially according to the execution order.

[0013] To achieve the above objectives, according to another aspect of this application, a business process operation and maintenance device is provided. The device includes: a response unit, configured to respond to a maintenance instruction for a target business process input by a user, and generate text prompts based on the original content of the maintenance instruction; a parsing unit, configured to perform semantic parsing on the text prompts, obtain parsing results, and decompose the text prompts into multiple maintenance operation information based on the parsing results; a generation unit, configured to generate corresponding maintenance task prompts based on each maintenance operation information, call a large language model to process the maintenance task prompts, and generate a policy maintenance instruction corresponding to each maintenance operation information; and a calling unit, configured to call the service interface indicated by each policy maintenance instruction to execute multiple policy maintenance operations.

[0014] In this embodiment, a maintenance instruction for the target business process is responded to by the user. Text prompts are generated based on the original content of the maintenance instruction. Semantic parsing of the text prompts yields the parsing results, which are then decomposed into multiple maintenance operation information. A corresponding maintenance task prompt is generated for each maintenance operation information. A large language model is called to process the maintenance task prompts, generating a policy maintenance instruction corresponding to each maintenance operation information. Multiple policy maintenance operations are executed by calling the service interface indicated by each policy maintenance instruction. By using a large language model to perform semantic parsing, decomposition, and task prompt generation of the original instruction, complex business maintenance instructions are automatically transformed into specific policy maintenance instructions. This replaces the tedious manual step-by-step configuration. Business personnel only need to input natural language instructions to complete multi-step policy maintenance without needing to understand the underlying interface details or complex configuration processes. This achieves the goal of improving policy maintenance efficiency, thereby improving the technical effect of improving business process operation and maintenance efficiency, and ultimately solving the technical problem of low business process operation and maintenance efficiency. Attached Figure Description

[0015] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings:

[0016] Figure 1 A hardware structure block diagram of a computer terminal (or mobile device) for implementing an operation and maintenance method for business processes is shown.

[0017] Figure 2 This is a flowchart of the operation and maintenance method for the business process provided in the embodiments of this application;

[0018] Figure 3 This is a schematic diagram of an operation and maintenance system based on the business process provided in the embodiments of this application;

[0019] Figure 4This is a schematic diagram of an operation and maintenance device for a business process provided in the embodiments of this application;

[0020] Figure 5 This is a structural block diagram of an electronic device according to an embodiment of this application. Detailed Implementation

[0021] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0022] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0023] It should be noted that the information collected in this application (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for display, data used for analysis, etc.) are 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 this data all comply with relevant laws, regulations, and standards, necessary confidentiality measures have been taken, and they do not violate public order and good morals. Corresponding access points are provided for users to choose to authorize or refuse. For example, interfaces are set up between this system and relevant users or organizations, providing users with corresponding access points to choose to agree to or refuse automated decision-making results; if the user chooses to refuse, the process proceeds to the expert decision-making stage.

[0024] Example 1

[0025] According to the embodiments of this application, a method embodiment for the operation and maintenance of a business process is also provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0026] The method embodiment provided in Embodiment 1 of this application can be executed on a mobile terminal, computer terminal, or similar computing device. Figure 1 A hardware structure block diagram of a computer terminal (or mobile device) for implementing an operation and maintenance method for business processes is shown. Figure 1 As shown, the computer terminal 10 (or mobile device) may include one or more processors 102 (shown as 102a, 102b, ..., 102n in the figure) 102 (processor 102 may include, but is not limited to, processing devices such as MCU (Microcontroller Unit) or FPGA (Field-Programmable Gate Array), memory 104 for storing data, and transmission device 106 for communication functions. In addition, it may also include: a display, input / output interfaces (I / O interfaces), a USB (Universal Serial Bus) port (which may be included as one of the ports of a BUS (Business Bus), a network interface, a power supply, and / or a camera. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the aforementioned electronic device. For example, computer terminal 10 may also include... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0027] It should be noted that the aforementioned one or more processors 102 and / or other data processing circuits are generally referred to herein as "data processing circuits". These data processing circuits may be embodied, in whole or in part, in software, hardware, firmware, or any other combination thereof. Furthermore, the data processing circuits may be a single, independent processing module, or may be integrated, in whole or in part, into any other element within the computer terminal 10 (or mobile device). As involved in the embodiments of this application, the data processing circuits serve as a processor control mechanism (e.g., selection of a variable resistor termination path connected to an interface).

[0028] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the operation and maintenance method of the business process in this embodiment. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, thereby realizing the above-mentioned operation and maintenance method of the business process. The memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to the computer terminal 10 via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0029] The transmission device 106 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by the communication provider of the computer terminal 10. In one example, the transmission device 106 includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device 106 may be a Radio Frequency (RF) module, used for wireless communication with the Internet.

[0030] The display may be, for example, a touchscreen LCD display that allows the user to interact with the user interface of the computer terminal 10 (or mobile device).

[0031] Under the above operating environment, this application provides a method for the operation and maintenance of business processes. Figure 2 This is a flowchart of the operation and maintenance method of the business process provided in the embodiments of this application, such as... Figure 2 As shown, the method includes:

[0032] Step S201: Respond to the maintenance instruction for the target business process input by the user, and generate text prompts based on the original content of the maintenance instruction.

[0033] In step S201, the target business process can be a business process such as financial business processing and customer outreach. In-store customer outreach refers to reaching customers through various channels and methods when they conduct business through the financial institution's online mobile application or offline branch, in order to assist them in completing their business. Maintenance instructions can be input instructions entered by business personnel on the front-end human-computer interaction interface via text, images, or voice, for customer outreach strategies such as adding, modifying, or deleting strategies. Customer outreach strategies are strategies maintained by business personnel on the business platform management end for reaching customers, including but not limited to strategy objectives, business types, customer acquisition methods, outreach methods, outreach paths, outreach scripts, outreach timing, effective time range, and strategy priorities. Text prompts are standardized, structured language text formed after the multimodal processor converts images and voice into text and constructs a parsing intelligent agent for processing. By uniformly converting diverse natural language or modal inputs from users into a standard text format that can be understood by the large model, and by constructing a parsing intelligent agent to invoke backend environment resources, a data foundation is laid for subsequent semantic understanding and task decomposition, solving the problem of traditional systems having a single input format and requiring manual preprocessing.

[0034] Step S202: Semantic parsing of the text prompt words to obtain the parsing results, and decomposing the text prompt words into multiple maintenance operation information based on the parsing results.

[0035] In step S202, semantic parsing is the process by which the intent recognition agent uses the semantic analysis capabilities of a large model to deeply understand text prompts and extract core intents and key entities such as business types, customer acquisition methods, and reach paths. The parsing result is structured data containing intent tags and key business parameters. Maintenance operation information maps high-level business intents to specific operation units of underlying atomic service types supported by the business platform, such as customer acquisition interfaces and reach method configuration interfaces. Through intent recognition and task decomposition, complex policy maintenance requirements are transformed into independent atomic units that the system can execute, eliminating the tedious process of manually configuring policies step by step and providing clear task guidance for automated execution.

[0036] Step S203: Generate corresponding maintenance task prompt words based on each maintenance operation information, call the large language model to process the maintenance task prompt words, and generate strategy maintenance instructions corresponding to each maintenance operation information.

[0037] In step S203, the large language model is a deep learning model trained on massive amounts of data with a huge parameter scale, capable of handling multimodal tasks such as language and images. Maintenance task prompts refer to sub-prompts automatically generated by the task decomposition agent based on maintenance operation information and preset atomic service interface specifications, containing specific parameter values ​​and execution constraints; policy maintenance instructions are specific data or code instructions output by the large language model after processing the maintenance task prompts, conforming to a specific service interface call format.

[0038] Step S204: Call the service interface indicated by each policy maintenance instruction to perform multiple policy maintenance operations.

[0039] In step S204, the service interface is the atomic service interface defined in the business platform policy maintenance system; the policy maintenance operation is the data writing, updating or deletion action actually completed by calling the above service interface, and multiple policy maintenance commands can be called concurrently or serially.

[0040] The business process operation and maintenance method provided in this application responds to the maintenance instructions of the target business process input by the user, generates text prompts based on the original content of the maintenance instructions; performs semantic parsing on the text prompts to obtain the parsing results, and decomposes the text prompts into multiple maintenance operation information based on the parsing results; generates corresponding maintenance task prompts based on each maintenance operation information, calls a large language model to process the maintenance task prompts, and generates a policy maintenance instruction corresponding to each maintenance operation information; and calls the service interface indicated by each policy maintenance instruction to execute multiple policy maintenance operations. By performing semantic parsing, decomposition, and task prompt generation on the original instructions through the large language model, complex business maintenance instructions are automatically transformed into specific policy maintenance instructions, replacing tedious manual step-by-step configuration. Business personnel only need to input natural language instructions to complete multi-step policy maintenance without needing to understand the underlying interface details or complex configuration processes, thereby achieving the goal of improving policy maintenance efficiency and thus realizing the technical effect of improving the operation and maintenance efficiency of business processes, thereby solving the technical problem of low operation and maintenance efficiency of business processes.

[0041] Upon receiving a maintenance instruction, a text prompt is generated by combining historical session content. Optionally, in the operation and maintenance method of the business process provided in this application embodiment, generating a text prompt based on the original content of the maintenance instruction includes: converting the images or voice in the original content into text content to obtain the target text content; obtaining the user's historical session content, merging the historical session content with preset policy maintenance prompts to generate an initial context; and constructing a text prompt based on the initial context and the target text content.

[0042] In some embodiments, the original content is a multi-modal customer engagement strategy maintenance instruction, including text, images, or voice, input by business personnel at the front-end human-computer interaction interface. The target text content is standardized plain text information extracted and converted from unstructured images or speech after processing by a multi-modal processor such as an optical character recognition module or an automatic speech recognition module. Historical conversation content, associated with the current user ID or session ID, is loaded from the short-term memory module by a parsing agent to maintain the continuity of the dialogue. Preset strategy maintenance prompts are system-level instructions built into the system, including role settings such as customer engagement strategy experts, business rule constraints such as compliance requirements, and task frameworks. The initial context is an input set containing complete background information formed by concatenating and fusing historical conversation content with preset strategy maintenance prompts. The text prompts are the final complete input instruction sequence sent to the large model for intent recognition and task decomposition after further integrating the above initial context and target text content.

[0043] This embodiment constructs a text prompt word rich in business context and standardized data by multimodal transformation, historical memory loading, and system prompt word fusion, which solves the problem of difficult multimodal input parsing. By introducing historical conversations and preset rules, it enhances the compliance and coherence of the large model output, improves the accuracy of subsequent intent recognition and task decomposition from the data preparation level, and provides solid data support for the automated generation of high-quality policy maintenance instructions.

[0044] After obtaining the text prompt words, semantic analysis is performed on them. Optionally, in the operation and maintenance method of the business process provided in this application embodiment, the semantic analysis of the text prompt words and the resulting analysis include: calling a large language model to process the text prompt words and obtaining the strategy maintenance type, wherein the strategy maintenance type includes at least one of the following: addition, deletion and modification; extracting business parameters from the text prompt words, wherein the business parameters include at least one of the following: target customer group, financial product and time range; and determining the business parameters and strategy maintenance type as the analysis results.

[0045] In some embodiments, the strategy maintenance type is the core operation category determined by the intent recognition agent after classifying the text prompts using a large language model. Specifically, it covers three basic atomic intents: adding (creating a new customer reach strategy), deleting, removing an existing strategy, and modifying / updating existing strategy attributes. Utilizing the classification capabilities of the large model, the core action intent of the user is accurately identified from complex natural language descriptions. Business parameters are key entity information extracted from the semantic context of the text prompts, used to concretize the customer reach strategy. These include the target customer group, financial products, and timeframes such as the start and end dates of the strategy's effectiveness and the timing of the outreach. The parsing result is a structured data object output by the intent recognition agent, which contains both the determined strategy maintenance type and business parameters, serving as the input for the task decomposition agent.

[0046] This embodiment accurately identifies policy maintenance types and extracts key business parameters through a large language model, transforming unstructured natural language input into parsed results. This achieves automated and high-precision identification of user intent and ensures the complete extraction of core policy elements such as customer groups, products, and time. It provides an accurate and standardized data foundation for subsequent task decomposition and atomic service calls, avoiding policy configuration failures caused by incorrect intent identification or missing parameters, and improving the automation level and robustness of policy management.

[0047] After obtaining the parsing results, the text prompt words are decomposed into multiple maintenance operation information. Optionally, in the operation and maintenance method of the business process provided in the embodiments of this application, decomposing the text prompt words into multiple maintenance operation information based on the parsing results includes: determining multiple service interfaces that need to be processed based on the policy maintenance type; determining the operation type for each service interface based on business parameters; and determining each service interface and the operation type corresponding to the service interface as a maintenance operation information to obtain multiple maintenance operation information.

[0048] In some embodiments, a service interface is an interface defined in the business platform policy maintenance system for performing specific underlying atomic operations, such as adding a policy interface, modifying a policy interface, and deleting a policy interface. An operation type is the specific action logic or data population method required for a particular service interface. For example, for a customer acquisition interface, the operation type might be building customer acquisition rules; for a reach configuration interface, the operation type might be setting reach scripts.

[0049] This embodiment filters service interfaces based on policy maintenance types and determines specific operation types by combining business parameters. It accurately decomposes complex natural language maintenance instructions into a series of structured maintenance operation information containing interfaces and operation logic, realizing a standardized mapping from business intent to atomic service calls, and ensuring the accuracy and completeness of task decomposition.

[0050] To ensure the accuracy of the target business process, the optimization completion of the target business process is determined by strategy execution indicators. Optionally, in the business process operation and maintenance method provided in this application embodiment, after executing multiple strategy maintenance operations by calling the service interface indicated by each strategy maintenance instruction, the method further includes: obtaining the strategy execution indicator value of the target business process within the target period, and determining whether the strategy execution indicator value is greater than or equal to the indicator value threshold; if the strategy execution indicator value is greater than or equal to the indicator value threshold, issuing a prompt message, wherein the prompt message is used to indicate that the target business process maintenance is complete; if the strategy execution indicator value is less than the indicator value threshold, optimizing the target business process in the strategy knowledge base, wherein the strategy knowledge base contains multiple business processes.

[0051] In some embodiments, the target period is a time range set by the system for evaluating the effectiveness of the strategy, such as 24 hours or 7 days after the strategy takes effect; the strategy execution indicator value is the quantitative performance data of the strategy within the period, such as click-through rate, conversion rate, and reach success rate, which is downloaded or queried in real time from the business platform's performance statistics system; the indicator value threshold is the performance qualification line preset by the system; the prompt information is the success status notification or confirmation message returned by the system to the front-end human-computer interaction interface; optimization is the optimization agent's use of the current inefficient strategy's execution data, combined with the retrieval enhancer to retrieve similar or historical successful cases from the strategy knowledge base, and using the large model to generate improvement suggestions; the "strategy knowledge base" stores a structured data set of existing customer reach strategies and their execution effect data, which is used to provide contextual support and decision-making basis for the large model.

[0052] This embodiment monitors policy execution metrics in real time and compares them with thresholds. This not only notifies users of maintenance completion when the policy is effective, improving interaction efficiency, but also automatically triggers an intelligent optimization process based on the policy knowledge base when the policy is ineffective, thus improving the level of intelligence in policy management.

[0053] Optimizing target business processes using historical reference strategies: Optionally, in the business process operation and maintenance method provided in this application embodiment, optimizing the target business process in the strategy knowledge base includes: retrieving historical reference strategies from the strategy knowledge base, wherein the similarity between the business scenario of the historical reference strategy and the business scenario of the target business process is greater than or equal to a preset similarity threshold; inputting the historical reference strategies and strategy execution indicator values ​​into a large language model, processing to obtain the strategy parameters to be adjusted and the recommended parameter values; setting the strategy parameters to be adjusted of the target business process to the recommended parameter values, thereby obtaining the optimized target business process.

[0054] In some embodiments, historical reference strategies are past successful cases or similar strategies retrieved from a knowledge base to assist in optimizing the current strategy; business scenarios are specific combinations of business conditions involved in the strategy, such as specific financial products, target customer characteristics, and outreach channels; preset similarity thresholds are used as criteria to determine which historical strategies are sufficiently comparable to the current inefficient strategy. The strategy parameters to be adjusted are the original strategy attributes that need to be modified, such as the original customer acquisition rules or scripts; recommended parameter values ​​are specific improvement suggestions or text generated by the large language model after integrating the characteristics of historical successful strategies with current failure data, such as adjusting the customer acquisition age range from 20-30 years old to 25-35 years old, or replacing it with a high-conversion outreach script.

[0055] This embodiment retrieves similar historical successful strategies and combines them with current failure data. It then uses a large model to intelligently generate specific parameter adjustment suggestions and automatically executes updates. This achieves data-driven adaptive optimization of business processes, improves the accuracy of parameter adjustments, and enhances the efficiency of business process operation and maintenance.

[0056] Policy maintenance operations need to be executed sequentially in the order of execution. Optionally, in the operation and maintenance method of the business process provided in the embodiments of this application, after decomposing the text prompt words into multiple maintenance operation information based on the parsing results, the method further includes: determining the execution order among multiple policy maintenance operations; calling the service interface indicated by each policy maintenance instruction to execute multiple policy maintenance operations includes: executing multiple policy maintenance operations sequentially in the order of execution.

[0057] In some embodiments, the execution order is determined by the logical dependencies between multiple maintenance operations, such as the need to complete customer acquisition to generate a target customer list before subsequent outreach operations can be performed; or the need to delete old policy configurations before modifying a policy, etc., which determine the sequential or parallel execution order of atomic tasks. The system scheduling module or intelligent agent collaboration mechanism analyzes the dependency graph between tasks to generate an execution sequence that conforms to business logic. Following the determined execution order, the service interfaces pointed to by each policy maintenance instruction are activated and called one by one, and the execution mechanism of the next operation is triggered only after the previous operation successfully returns a result.

[0058] This embodiment solves the timing conflicts and data inconsistencies that may be caused by multi-step operations in complex policy maintenance scenarios by intelligently determining the logical dependency order between multiple atomic maintenance operations and strictly executing them in sequence. This ensures the rigor and integrity of the policy configuration process, thereby guaranteeing the high availability and execution success rate of the automated policy generation and update mechanism in actual business environments.

[0059] According to another embodiment of this application, a business process operation and maintenance system is also provided. Figure 3This is a schematic diagram of an operation and maintenance system based on the business process provided in the embodiments of this application, such as... Figure 3 As shown, the system consists of a front-end human-computer interaction interface, a large model back-end processing node, and a strategy maintenance system and effect statistics system for the business platform.

[0060] Specifically, the front-end human-computer interaction interface only interacts with the back-end processing node of the large model. Business personnel only need to enter the customer engagement strategy they want to maintain in the input box of the front-end human-computer interaction interface, and the front-end interaction will send the content to the back-end processing node of the large model to complete the parsing and processing of the input content.

[0061] The large model's backend processing nodes receive, parse, and process the customer engagement strategy content sent from the frontend interaction, decomposing it into multiple atomic tasks. These tasks call the atomic services of the strategy maintenance system in the business platform to add, modify, and delete customer engagement strategies. On the other hand, the backend processing nodes download statistical data of customer engagement strategies from the performance statistics system in the business platform in real time, analyze the strategy's effectiveness, optimize the customer engagement strategies based on the analysis data, and continuously adjust the strategies to improve customer engagement results.

[0062] The backend processing node architecture of the large model consists of the following:

[0063] Asynchronous event-driven: Primarily addresses latency issues in large models, employing message queues to achieve streaming responses and facilitate data interaction between the front-end and the back-end processing nodes of the large model; Multimodal processor: Extracts images, audio, etc., into text, serving as input prompts for the large model; Variable processor: Extracts strategy parameters from input, passes them to the strategy maintenance system, stores important parameters, and dynamically adjusts them in conjunction with the strategy analysis system; Retrieval enhancement generator: Builds a strategy knowledge base based on existing customer engagement strategies, providing data support for dynamic analysis and optimization of strategies;

[0064] Multi-agent collaboration: This core system includes a construction and parsing agent, a task decomposition agent, an intent recognition agent, an execution agent, and an optimization agent. The construction and parsing agent primarily handles input prompts and mobilizes backend resources of the large model. The task decomposition agent breaks down input prompts into multiple atomic prompts needed for each service and automatically concatenates and optimizes them. The intent recognition agent performs intent recognition on the decomposed task prompts and provides them as input to the execution agent. The execution agent, after receiving the decomposed atomic prompts, executes its respective tasks, such as adding, deleting, and modifying customer engagement strategies. The optimization agent analyzes and optimizes based on a strategy knowledge base and dynamically adjusts strategy parameters through a variable processor. These multiple agents are interdependent and collaborate collaboratively.

[0065] The business process operation and maintenance system in this embodiment does not require manual analysis of customer reach effectiveness. It can quickly obtain analysis data using large model technology, thereby optimizing customer reach strategies. At the same time, business personnel can still log in to the business platform to view the maintained strategy information, which improves the efficiency of business process operation and maintenance.

[0066] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0067] Example 2

[0068] This application also provides a business process operation and maintenance device. It should be noted that this business process operation and maintenance device can be used to execute the business process operation and maintenance method provided in this application. The following describes the business process operation and maintenance device provided in this application.

[0069] According to an embodiment of this application, an apparatus for implementing the above-described business process operation and maintenance method is also provided. Figure 4 This is a schematic diagram of an operation and maintenance device for a business process provided in an embodiment of this application, such as... Figure 4 As shown, the device includes:

[0070] The response unit 401 is used to respond to the maintenance instructions of the target business process input by the user and generate text prompts based on the original content of the maintenance instructions.

[0071] The parsing unit 402 is used to perform semantic parsing on the text prompt words, obtain the parsing results, and decompose the text prompt words into multiple maintenance operation information based on the parsing results;

[0072] The generation unit 403 is used to generate corresponding maintenance task prompt words based on each maintenance operation information, call the large language model to process the maintenance task prompt words, and generate strategy maintenance instructions corresponding to each maintenance operation information.

[0073] Calling unit 404 is used to call the service interface indicated by each policy maintenance instruction to perform multiple policy maintenance operations.

[0074] The business process operation and maintenance device provided in this application embodiment responds to the maintenance instructions of the target business process input by the user through a response unit 401, and generates text prompts based on the original content of the maintenance instructions; a parsing unit 402 performs semantic parsing on the text prompts to obtain the parsing results, and decomposes the text prompts into multiple maintenance operation information based on the parsing results; a generation unit 403 generates corresponding maintenance task prompts based on each maintenance operation information, calls a large language model to process the maintenance task prompts, and generates a policy maintenance instruction corresponding to each maintenance operation information; and a calling unit 404 calls the service interface indicated by each policy maintenance instruction to execute multiple policy maintenance operations. By performing semantic parsing, decomposition, and task prompt generation on the original instructions through a large language model, complex business maintenance instructions are automatically transformed into specific policy maintenance instructions, replacing tedious manual step-by-step configuration. Business personnel only need to input natural language instructions to complete multi-step policy maintenance without needing to understand the underlying interface details or complex configuration processes, thereby achieving the goal of improving policy maintenance efficiency and thus realizing the technical effect of improving the efficiency of business process operation and maintenance, and solving the technical problem of low efficiency in business process operation and maintenance.

[0075] Optionally, in the operation and maintenance device for the business process provided in this application embodiment, the response unit 401 includes: a conversion module, used to convert the images or voice in the original content into text content to obtain the target text content; an acquisition module, used to acquire the user's historical session content, merge the historical session content with preset policy maintenance prompt words, and generate an initial context; and a construction module, used to construct text prompt words based on the initial context and the target text content.

[0076] Optionally, in the business process operation and maintenance device provided in this application embodiment, the parsing unit 402 includes: a calling module, used to call a large language model to process text prompt words and obtain a strategy maintenance type, wherein the strategy maintenance type includes at least one of the following: addition, deletion, and modification; an extraction module, used to extract business parameters from the text prompt words, wherein the business parameters include at least one of the following: target customer group, financial product, and time range; and a first determining module, used to determine the business parameters and strategy maintenance type as the parsing result.

[0077] Optionally, in the business process operation and maintenance device provided in the embodiments of this application, the parsing unit 402 includes: a second determining module, used to determine multiple service interfaces that need to be processed based on the policy maintenance type; a third determining module, used to determine the operation type for each service interface based on business parameters; and a fourth determining module, used to determine each service interface and the operation type corresponding to the service interface as a maintenance operation information, thereby obtaining multiple maintenance operation information.

[0078] Optionally, in the business process operation and maintenance device provided in the embodiments of this application, the device further includes: an acquisition unit, used to acquire the strategy execution indicator value of the target business process within a target period, and determine whether the strategy execution indicator value is greater than or equal to the indicator value threshold; a prompting unit, used to issue a prompting message when the strategy execution indicator value is greater than or equal to the indicator value threshold, wherein the prompting message is used to indicate that the maintenance of the target business process is completed; and an optimization unit, used to optimize the target business process in the strategy knowledge base when the strategy execution indicator value is less than the indicator value threshold, wherein the strategy knowledge base contains multiple business processes.

[0079] Optionally, in the operation and maintenance device for the business process provided in this application embodiment, the optimization unit includes: a retrieval module, used to retrieve historical reference strategies from the strategy knowledge base, wherein the similarity between the business scenario of the historical reference strategy and the business scenario of the target business process is greater than or equal to a preset similarity threshold; an input module, used to input the historical reference strategy and strategy execution indicator values ​​into a large language model, and process them to obtain the strategy parameters to be adjusted and the recommended parameter values; and a setting module, used to set the strategy parameters to be adjusted of the target business process to the recommended parameter values, thereby obtaining the optimized target business process.

[0080] Optionally, in the business process operation and maintenance device provided in the embodiments of this application, the device further includes: a determining unit, used to determine the execution order among multiple policy maintenance operations; the calling unit 404 includes: an execution module, used to execute multiple policy maintenance operations sequentially according to the execution order.

[0081] It should be noted that the response unit 401, parsing unit 402, generation unit 403, and calling unit 404 mentioned above correspond to steps S201 to S204 in Embodiment 1. The four units and the corresponding steps implement the same instances and application scenarios, but are not limited to the content disclosed in Embodiment 1. It should be noted that the above modules or units can be hardware or software components stored in memory (e.g., memory 104) and processed by one or more processors (e.g., processors 102a, 102b, ..., 102n). The above modules or units can also be part of a device and run in the computer terminal 10 provided in Embodiment 1.

[0082] Example 3

[0083] Embodiments of this application may provide an electronic device. Figure 5 This is a structural block diagram of an electronic device according to an embodiment of this application. Figure 5 As shown, the electronic device may include: one or more ( Figure 5(Only one is shown) processor 502, memory 504, memory controller, and peripheral interface, wherein the peripheral interface is connected to the radio frequency module, audio module and display.

[0084] The memory can be used to store software programs and modules, such as the program instructions / modules corresponding to the methods and apparatus in the embodiments of this application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, thereby implementing the above-described methods. The memory may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include memory remotely located relative to the processor, and these remote memories can be connected to the terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0085] The processor can access information and applications stored in memory via a transmission device to perform the following steps: responding to maintenance instructions for the target business process input by the user, generating text prompts based on the original content of the maintenance instructions; performing semantic parsing on the text prompts to obtain the parsing results, and decomposing the text prompts into multiple maintenance operation information based on the parsing results; generating corresponding maintenance task prompts based on each maintenance operation information, calling a large language model to process the maintenance task prompts, and generating policy maintenance instructions corresponding to each maintenance operation information; and calling the service interface indicated by each policy maintenance instruction to execute multiple policy maintenance operations.

[0086] The processor can also access information and applications stored in the memory via a transmission device to perform the following steps: converting images or speech in the original content into text content to obtain the target text content; acquiring the user's historical conversation content, merging the historical conversation content with preset policy maintenance prompts to generate an initial context; and constructing text prompts based on the initial context and the target text content.

[0087] The processor can also invoke information and applications stored in the memory via the transmission device to perform the following steps: invoke the large language model to process the text prompt words to obtain the strategy maintenance type, wherein the strategy maintenance type includes at least one of the following: addition, deletion and modification; extract business parameters from the text prompt words, wherein the business parameters include at least one of the following: target customer group, financial product and time range; determine the business parameters and strategy maintenance type as the parsing result.

[0088] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: determine multiple service interfaces that need to be processed based on the policy maintenance type; determine the operation type for each service interface based on the business parameters; determine each service interface and the operation type corresponding to the service interface as a maintenance operation information to obtain multiple maintenance operation information.

[0089] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: obtain the policy execution indicator value of the target business process within the target period, and determine whether the policy execution indicator value is greater than or equal to the indicator value threshold; if the policy execution indicator value is greater than or equal to the indicator value threshold, issue a prompt message, wherein the prompt message is used to indicate that the maintenance of the target business process is completed; if the policy execution indicator value is less than the indicator value threshold, optimize the target business process in the policy knowledge base, wherein the policy knowledge base contains multiple business processes.

[0090] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: retrieve historical reference strategies from the strategy knowledge base, wherein the similarity between the business scenario of the historical reference strategy and the business scenario of the target business process is greater than or equal to a preset similarity threshold; input the historical reference strategies and strategy execution index values ​​into the large language model, process them to obtain the strategy parameters to be adjusted and the recommended parameter values; set the strategy parameters to be adjusted of the target business process to the recommended parameter values ​​to obtain the optimized target business process.

[0091] The processor can also invoke information and applications stored in memory via a transmission device to perform the following steps: determining the execution order among multiple policy maintenance operations; invoking the service interface indicated by each policy maintenance instruction to perform multiple policy maintenance operations, including: performing multiple policy maintenance operations sequentially in the execution order.

[0092] This application provides a scheme for responding to user input of maintenance instructions for a target business process. The scheme generates text prompts based on the original content of the maintenance instructions; performs semantic parsing on the text prompts to obtain the parsing results; decomposes the text prompts into multiple maintenance operation information based on the parsing results; generates corresponding maintenance task prompts for each maintenance operation information; calls a large language model to process the maintenance task prompts to generate a policy maintenance instruction corresponding to each maintenance operation information; and calls the service interface indicated by each policy maintenance instruction to execute multiple policy maintenance operations. By using a large language model to perform semantic parsing, decomposition, and task prompt generation on the original instructions, complex business maintenance instructions are automatically transformed into specific policy maintenance instructions, replacing tedious manual step-by-step configuration. Business personnel only need to input natural language instructions to complete multi-step policy maintenance without needing to understand the underlying interface details or complex configuration processes, thus improving the efficiency of policy maintenance and achieving the technical effect of improving the operational efficiency of business processes, thereby solving the technical problem of low operational efficiency of business processes.

[0093] Those skilled in the art will understand that Figure 5 The structure shown is for illustrative purposes only. Electronic devices can also be smartphones, tablets, handheld computers, mobile internet devices (MIDs), PADs, and other terminal devices. Figure 5 This does not limit the structure of the aforementioned electronic device. For example, electronic devices may also include components that are more... Figure 5 The more or fewer components shown (such as network interfaces, display devices, etc.), or having the same Figure 5 The different configurations shown.

[0094] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing the hardware related to the terminal device. The program can be stored in a computer-readable storage medium, which may include: flash drive, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0095] Example 4

[0096] Embodiments of this application also provide a storage medium. Optionally, in this embodiment, the storage medium can be used to store the program code executed by the operation and maintenance method of the business process provided in Embodiment 1.

[0097] Optionally, in this embodiment, the storage medium may be located in any computer terminal in a group of computer terminals in a computer network, or in any mobile terminal in a group of mobile terminals.

[0098] This application also provides a computer program product, which, when executed on a data processing device, is suitable for performing operation and maintenance method steps of a business process.

[0099] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0100] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0101] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0102] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0103] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0104] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard drive, magnetic disk, or optical disk.

[0105] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for operating and maintaining a business process, characterized in that, include: Responding to the maintenance instructions for the target business process input by the user, generating text prompts based on the original content of the maintenance instructions; The text prompt words are semantically parsed to obtain the parsing results, and the text prompt words are decomposed into multiple maintenance operation information based on the parsing results; Based on each maintenance operation information, a corresponding maintenance task prompt word is generated, and a large language model is called to process the maintenance task prompt word to generate a strategy maintenance instruction corresponding to each maintenance operation information. Each policy maintenance instruction indicates a service interface to perform multiple policy maintenance operations.

2. The method according to claim 1, characterized in that, The text prompts generated based on the original content of the maintenance instructions include: The images or audio in the original content are converted into text content to obtain the target text content; Obtain the user's historical conversation content, and merge the historical conversation content with preset policy maintenance prompts to generate an initial context; The text prompt word is constructed based on the initial context and the target text content.

3. The method according to claim 1, characterized in that, Semantic analysis of the text prompt words yields the following results: The large language model is invoked to process the text prompt words to obtain the strategy maintenance type, wherein the strategy maintenance type includes at least one of the following: addition, deletion, and modification; Business parameters are extracted from the text prompts, wherein the business parameters include at least one of the following: target customer group, financial product, and time range; The business parameters and the strategy maintenance type are determined as the parsing result.

4. The method according to claim 3, characterized in that, Based on the parsing results, the text prompt words are decomposed into multiple maintenance operation information, including: Based on the strategy maintenance type, determine the multiple service interfaces that need to be processed; The operation type for each service interface is determined based on the business parameters. Each service interface and its corresponding operation type are defined as a maintenance operation information, thus obtaining the plurality of maintenance operation information.

5. The method according to claim 1, characterized in that, After performing multiple policy maintenance operations by invoking the service interface indicated by each of the policy maintenance instructions, the method further includes: Obtain the strategy execution indicator value of the target business process within the target period, and determine whether the strategy execution indicator value is greater than or equal to the indicator value threshold; If the strategy execution indicator value is greater than or equal to the indicator value threshold, a prompt message is issued, wherein the prompt message is used to indicate that the target business process maintenance is complete; If the strategy execution indicator value is less than the indicator value threshold, optimize the target business process in the strategy knowledge base, wherein the strategy knowledge base contains multiple business processes.

6. The method according to claim 5, characterized in that, The target business processes mentioned in the optimization strategy knowledge base include: Historical reference strategies are retrieved from the strategy knowledge base, wherein the similarity between the business scenario of the historical reference strategy and the business scenario of the target business process is greater than or equal to a preset similarity threshold. The historical reference strategy and the strategy execution index value are input into the large language model to obtain the strategy parameters to be adjusted and the recommended parameter values. The optimized target business process is obtained by setting the strategy parameters to be adjusted in the target business process to the recommended parameter values.

7. The method according to claim 1, characterized in that, After decomposing the text prompt words into multiple maintenance operation information based on the parsing results, the method further includes: Determine the execution order among the multiple strategy maintenance operations; Invoking the service interface indicated by each of the aforementioned policy maintenance instructions to perform multiple policy maintenance operations includes: The multiple policy maintenance operations are executed sequentially according to the execution order.

8. A business process operation and maintenance device, characterized in that, include: A response unit is used to respond to maintenance instructions for the target business process input by the user, and to generate text prompts based on the original content of the maintenance instructions; The parsing unit is used to perform semantic parsing on the text prompt words, obtain the parsing result, and decompose the text prompt words into multiple maintenance operation information based on the parsing result; The generation unit is used to generate a corresponding maintenance task prompt word based on each maintenance operation information, call a large language model to process the maintenance task prompt word, and generate a strategy maintenance instruction corresponding to each maintenance operation information. The invocation unit is used to invoke the service interface indicated by each of the policy maintenance instructions to perform multiple policy maintenance operations.

9. An electronic device, characterized in that, include: Memory, which stores executable programs; A processor for running the program, wherein the program, when running, executes the operation and maintenance method of the business process according to any one of claims 1 to 7.

10. A computer program product comprising computer instructions, characterized in that, When the computer instructions are executed by the processor, they implement the steps of the operation and maintenance method of the business process as described in any one of claims 1 to 7.

Images