Application construction method and computing device

Abstract

The embodiment of the invention provides an application construction method and computing equipment. The method comprises the steps of obtaining an initial workflow in response to a modification request for a target application; the initial workflow is used for representing a workflow of the target application; obtaining a modification suggestion input by a user in the first-stage dialogue; the modification suggestion is used for describing an improvement direction for the initial workflow; adjusting the initial workflow into a target workflow by adopting a first intelligent agent; the target workflow conforms to the modification suggestion; and generating a target application based on the target workflow. The method at least can improve the efficiency of application construction and reduce the development cycle of the application.

Description

Technical Field

[0001] This application relates to the field of computing technology, and in particular to an application construction method and computing device. Background Technology

[0002] With the development of artificial intelligence technology, large models are being applied more and more widely in various fields. Large models can understand and generate natural language text, thereby assisting users in solving a wide variety of problems. While large models possess powerful general-purpose capabilities, for users in specialized fields, these general-purpose models often cannot efficiently solve business problems specific to their chosen domain. Therefore, large model applications have emerged. Large model applications are based on large models and integrate specialized knowledge bases, thus effectively assisting users in solving business problems within specific professional domains.

[0003] However, current construction solutions for large-scale model applications are quite complex and require many steps, which can affect the efficiency of application construction. Summary of the Invention

[0004] This application provides an application building method and computing device, which can at least improve the efficiency of application building and reduce the application development cycle.

[0005] In a first aspect, the application building method provided in this application includes: responding to a modification request for a target application and obtaining an initial workflow; the initial workflow is used to represent the workflow of the target application;

[0006] Obtain modification suggestions input by the user during the first phase of the dialogue; the modification suggestions are used to describe directions for improvement for the initial workflow;

[0007] The initial workflow is adjusted using a first intelligent agent to obtain a candidate workflow; the candidate workflow includes a target node; the target node is a node in the initial workflow that has been modified based on the modification suggestion, or a node in the initial workflow that has been added based on the modification suggestion.

[0008] The target workflow is obtained by configuring the function implementation scheme of the target node using a second intelligent agent; the target workflow conforms to the modification suggestion.

[0009] Based on the target workflow, the target application is generated.

[0010] In the application building method provided in this application embodiment, the computing device obtains an initial workflow and acquires modification suggestions from the user through a first-stage dialogue. Further, a first intelligent agent is used to process the initial workflow to obtain candidate workflows. Further, a second intelligent agent is used to configure the functional implementation scheme of the target nodes in the candidate workflows to obtain a target workflow that meets the modification suggestions, and then a target application is generated based on the target workflow. It can be seen that during application development and modification, users only need to input modification opinions, making the building process simple. The solution of this application can adjust the workflow flow according to the user's modification suggestions through the first intelligent agent and configure the functional implementation scheme of the workflow through the second intelligent agent, ultimately obtaining a target workflow that meets user needs and can run. Through the above method, the efficiency of application modification is effectively improved, and the complexity and possibility of errors in manual modification operations by users are reduced.

[0011] In one possible implementation, the step of using a second agent to configure the functional implementation scheme of the target node to obtain the target workflow includes: using the second agent to determine a node template that matches the modification suggestion from a set of node templates, and using the node template to configure the functional implementation scheme of the target node; the set of node templates includes multiple node templates, and each node template includes a functional implementation scheme for one function.

[0012] In another possible implementation, if no node template matching the modification suggestion exists in the set of node templates, the method further includes: using a third agent to generate a function implementation scheme based on the function description of the target node, and configuring the target node using the function implementation scheme.

[0013] In another possible implementation, the method further includes: using a fourth agent to verify the target workflow to determine whether the target workflow is running normally; and generating the target application based on the target workflow, including: generating the target application based on the target workflow if it is determined that the target workflow is running normally.

[0014] In another possible implementation, the step of using a fourth intelligent agent to verify the target workflow includes: obtaining initial input information of the target workflow; using the fourth intelligent agent to input the initial input information into the target workflow; and the fourth intelligent agent sequentially debugging each node according to the direction of the target workflow.

[0015] In another possible implementation, the method further includes: when the target workflow cannot operate normally, obtaining a fault problem determined by a fourth agent; the fault problem is used to indicate the fault node and the cause of the fault; the first agent adjusts the target workflow based on the fault problem, and the fourth agent verifies the adjusted workflow until the target workflow operates normally.

[0016] In another possible implementation, the method further includes: using a first intelligent agent to generate a modification scheme based on the modification suggestion, and displaying the modification scheme through a display device; if the user rejects the modification scheme, the first intelligent agent regenerates the modification scheme; using the first intelligent agent to adjust the initial workflow to the target workflow, including: if the user confirms the modification scheme, using the first intelligent agent to adjust the initial workflow to the target workflow based on the modification scheme.

[0017] In another possible implementation, obtaining the modification suggestions input by the user in the first stage of the dialogue includes: obtaining the overall goal of the target application and the modification suggestions; if the modification suggestions do not conform to the overall goal, prompting the user to re-enter the modification suggestions.

[0018] In another possible implementation, generating the target application based on the target workflow includes: obtaining attribute information of the target application; the attribute information includes at least one of the following: name, version number, icon, online status; and using a fifth intelligent agent to generate the target application based on the target workflow and the attribute information.

[0019] Secondly, this application provides an application building apparatus for executing any of the application building methods provided in the first aspect above.

[0020] Thirdly, embodiments of this application provide a computing device including a processor and a memory; the processor is coupled to the memory; the memory is used to store computer instructions, which are loaded and executed by the processor to enable the computing device to implement the method described in the first aspect.

[0021] Fourthly, embodiments of this application provide a computer-readable storage medium comprising: computer software instructions; when the computer software instructions are executed in a computing device, they cause the computing device to implement the method described in the first aspect.

[0022] Fifthly, embodiments of this application provide a computer program product that, when run on a computing device, causes the computing device to execute the steps of the relevant method described in the first aspect above, so as to implement the method of the first aspect above.

[0023] The beneficial effects of the second to fifth aspects mentioned above can be referred to the corresponding description of the first aspect, and will not be repeated here. Attached Figure Description

[0024] Figure 1 This application provides an architecture diagram of an application building system.

[0025] Figure 2 This is a schematic diagram of an application scenario provided by an embodiment of this application;

[0026] Figure 3 This is a schematic diagram of the hardware structure of a computing device provided in an embodiment of this application;

[0027] Figure 4 A flowchart illustrating an application construction method provided in an embodiment of this application;

[0028] Figure 5 A schematic diagram of a workflow provided for an embodiment of this application;

[0029] Figure 6 A flowchart illustrating a workflow generation method provided in this application embodiment;

[0030] Figure 7 A flowchart illustrating a workflow adjustment provided in an embodiment of this application;

[0031] Figure 8 A schematic diagram illustrating a specific example provided in an embodiment of this application;

[0032] Figure 9 A schematic diagram of an application building apparatus provided in an embodiment of this application;

[0033] Figure 10 This is a schematic diagram of the composition of a computing device provided in an embodiment of this application. Detailed Implementation

[0034] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0035] It should be noted that in the embodiments of this application, the words "exemplarily" or "for example" are used to indicate examples, illustrations, or explanations. Any embodiment or design scheme described as "exemplarily" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design schemes. Specifically, the use of the words "exemplarily" or "for example" is intended to present the relevant concepts in a specific manner.

[0036] To facilitate a clear description of the technical solutions of the embodiments of this application, the terms "first" and "second" are used in the embodiments of this application to distinguish the same or similar items with essentially the same function and effect. Those skilled in the art can understand that the terms "first" and "second" are not intended to limit the quantity or execution order.

[0037] The following is a brief explanation of the technical terms used in the embodiments of this application:

[0038] 1. Large Language Model (LLM): A type of artificial intelligence model that uses deep learning algorithms to learn language patterns and structures through training on a large amount of text data, thereby enabling it to understand and generate natural language text.

[0039] 2. Intelligent Agent: This refers to an intelligent system built upon a large model, capable of autonomously understanding tasks, planning steps, invoking tools, and achieving complex goals. Through interaction with the large model, it combines natural language understanding, reasoning abilities, and external tools to achieve problem-solving capabilities similar to humans.

[0040] 3. Enterprise knowledge base: An internal system or platform used by an enterprise to store, manage and share knowledge resources. It stores various documents of the enterprise, such as policies, processes, manuals, training materials, etc., and records the experience and lessons learned by the enterprise in different projects and tasks, which are used to help employees solve problems encountered in their work.

[0041] 4. Prompt: This refers to the input text used to guide or stimulate the large language model to perform a specific task. Its purpose is to help the large language module model understand the type of task the user wants to perform or the required output format.

[0042] 5. Retrieval-augmented generation (RAG): This technology allows users to retrieve answers relevant to their questions from a pre-configured enterprise knowledge vector database. These answers, after being concatenated with prompts, are then fed into a larger model to help it better answer user questions.

[0043] The current approach to building applications using large models involves providing a simple interface through which users input their application requirements. The computing device then uses the large model to automatically select the necessary components (code, database, third-party tools, etc.) based on these requirements, ultimately generating the application. However, this approach often results in issues after application creation, requiring further adjustments. The current adjustment process is quite complex and can impact application build efficiency.

[0044] Based on this, this application provides an application building method that uses an intelligent agent to adjust the application workflow based on user input modification suggestions, thereby improving the application modification efficiency and shortening the application development cycle.

[0045] In some implementations, the computing device acquires an initial workflow and obtains user-inputted modification suggestions through a first-stage dialogue. Further, a first intelligent agent processes the initial workflow to obtain candidate workflows. Further still, a second intelligent agent configures the functional implementation schemes of target nodes in the candidate workflows to obtain a target workflow that meets the modification suggestions, and then generates a target application based on the target workflow. It can be seen that during application development and modification, users only need to input their modification opinions, simplifying the construction process. The solution of this application can adjust the workflow flow based on user modification suggestions through a first intelligent agent and configure the functional implementation schemes of the workflow through a second intelligent agent, ultimately obtaining a target workflow that meets user needs and can run. Through the above methods, the efficiency of application modification is effectively improved, and the complexity and possibility of errors in manual modification operations by users are reduced.

[0046] The embodiments provided in this application will now be described in detail with reference to the accompanying drawings.

[0047] Figure 1 This is a schematic diagram of the architecture of an application building system provided in an embodiment of this application. The system achieves a fully automated process from user requirement collection to application generation through the collaborative work of multiple modules, and relies on global variables to achieve data sharing and state synchronization among the modules. Here, global variables refer to variables that can be accessed and used by every module in the application building system.

[0048] 1. System Architecture and Data Flow

[0049] The application build system operates on global variables, which store structured data generated by each module. This structured data is crucial for building the application, including:

[0050] The application name (application name = []), functional requirements (application function = []), sample data (example = []), and operation process description (application operation process = []) obtained by the information collection module are all user inputs.

[0051] The preset initial workflow includes nodes (nodes = []) and edges (edges = []), which together form the topology of the application logic.

[0052] The node configuration information (configs=[]) obtained by the configuration generation module includes the specific implementation scheme of each node in the workflow, which can support the operation of the application.

[0053] 2. Core Module Function Description

[0054] (1) Information gathering agent (hereinafter referred to as the seventh agent)

[0055] This agent is responsible for interacting with users using natural language, extracting key information from the user's input (such as user suggestions for application modifications, and a description of the application's overall goals). After each round of interaction, it checks whether the collected key information meets the application's construction conditions. If the conditions are met, the key information is stored in a global variable, triggering other agents to begin working. Here, the application construction conditions can be understood as the clarity, richness, and completeness of the key information. When the key information meets the clarity, richness, and completeness requirements for application construction, it is more conducive to building an application that meets the user's needs.

[0056] (2) Workflow adjustment agent (hereinafter referred to as the first agent)

[0057] It allows users to obtain modification suggestions for their input through natural language interaction and adjust the specified workflow to obtain a workflow that meets the user's needs.

[0058] (3) Node configuration agent (hereinafter referred to as the second agent)

[0059] The target node in the workflow (a newly added or modified node) lacks specific implementation details (functional implementation scheme). Therefore, the node configuration module retrieves a suitable functional implementation scheme from the node template set and configures the target node, thus defining the specific implementation details of the target node. The node template set includes multiple node templates, and each node template includes a functional implementation scheme for one function.

[0060] (4) Configure and generate intelligent agents (hereinafter referred to as third intelligent agents)

[0061] If no suitable implementation scheme for the target node can be found in the node template set, the configuration agent will generate the implementation scheme based on the target node.

[0062] (5) Workflow verification agent (hereinafter referred to as the fourth agent)

[0063] Perform node-by-node testing on the generated workflow to verify that the inputs / outputs meet expectations. During debugging, return the execution results of each node and the final running status of the entire workflow for user or system optimization reference.

[0064] (6) Application of Generated Intelligent Agents

[0065] Add metadata (such as application name, version, etc.) to the debugged workflow and package it into a deployable application.

[0066] It should be noted that, Figure 1 This is merely one example; the application building system may also include more or fewer agents, and this application embodiment does not specifically limit this.

[0067] Figure 2 A schematic diagram illustrating an application scenario provided by an embodiment of this application is shown. For example... Figure 2 As shown, it includes a terminal device 200 and a server 210. The application building method provided in this embodiment can be executed by the terminal device 200.

[0068] The terminal device 200 and the server 210 can interact via a communication network. This communication network can be Ethernet, wireless access network, wireless local area network (WLAN), etc.

[0069] Among them, terminal device 200 is a device with interface display function.

[0070] Optionally, the terminal device 200 can be a personal computer, smartphone, tablet computer, e-book reader, portable computer, or other similar device.

[0071] Server 210 deploys a large language model, providing an interactive interface for the large language model to terminal device 200, and allowing terminal device 200 to obtain descriptive information about the application the user needs to build. Then, server 210 uses the large language model to understand the descriptive information, constructs the application workflow, and displays it to the user through terminal device 210. The user can adjust the workflow according to their needs.

[0072] For example, terminal device 200 can access server 210 through a browser to display the interactive interface of the large language model.

[0073] Optionally, the aforementioned server 210 can be a cluster of one or more servers. Figure 1 Each intelligent agent in the process can be deployed on the same server or on different servers. This application does not specifically limit this.

[0074] In terms of form, the server mentioned here can be a blade server, a high-density server, a rack server, or a full-rack server; in terms of function, the server can be a general-purpose server, a graphics processing unit (GPU) server, an artificial intelligence (AI) server, etc.

[0075] Figure 3 This is a schematic diagram of the hardware structure of a computing device (which may be the server 210 described above) provided in an embodiment of this application. Figure 3 As shown, the computing device may include a processor and a memory; the memory is used to store computer instructions, which are loaded and executed by the processor, enabling the computing device to implement the application building method described in the following embodiments.

[0076] It should be noted that the system architecture and application scenarios described in the embodiments of this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided in the embodiments of this application. As those skilled in the art will know, with the evolution of system architecture and the emergence of new business scenarios, the technical solutions provided in the embodiments of this application are also applicable to similar technical problems.

[0077] Figure 4 This is a flowchart illustrating an application building method provided in an embodiment of this application. For example, the application building method provided in this embodiment can be applied to... Figure 3 In the computing device shown.

[0078] like Figure 4 As shown, the application building method provided in this application embodiment may include the following steps:

[0079] S401. In response to a modification request for the target application, obtain the initial workflow.

[0080] The initial workflow is used to represent the workflow of the target application.

[0081] In this embodiment, when a user needs to build an application, they initiate a modification request on the computing device to start the application build process. The computing device can then obtain the initial workflow that needs modification.

[0082] In some embodiments, the initial workflow can be a workflow input by the user or a workflow generated temporarily by the workflow generating agent based on user needs.

[0083] The workflow is explained in detail below.

[0084] Workflow, also known as a diagram, represents an application in a single diagram, such as... Figure 5 As shown, a workflow consists of nodes and edges. Nodes are responsible for executing specific tasks, while edges represent the flow of those tasks. Users can determine the overall workflow of the application, including what nodes and edges are needed and the order in which they are connected.

[0085] It's important to note that while each node corresponds to a specific task, some tasks have relatively fixed functionalities and cannot be implemented using a single functional node. For example, Application Programming Interface (API) calls not only require code support to implement the call, but also require the user to manually pass parameters during the call process. In other words, although API calls involve multiple steps, these steps are generally a fixed combination. Therefore, multiple steps can be combined into one step and displayed as a node in the workflow to facilitate workflow construction.

[0086] It's important to note that each node in the application's workflow includes two levels of attributes: a functional description and a functional implementation plan. For example, if the functional description of a node is to translate input Chinese into English, the corresponding implementation plan would be a series of codes that import the input Chinese into the translator and obtain the translation result.

[0087] For a single node, its functionality can be implemented in the following ways: for example, a combination of prompt words and a large model; it can also retrieve results from a specified database using RAG technology; or it can utilize third-party tools (such as translation tools provided by other companies). Therefore, a specific implementation of a node can be derived from one or more combinations of prompt words, large models, codes, RAG, and tools.

[0088] Furthermore, nodes in a workflow can be categorized into functional nodes that include decision-making capabilities and functional nodes that do not. Functional nodes that include decision-making capabilities can be called logical nodes (decision nodes). Logical nodes are used to establish logical connections between functional nodes; for example, based on the output of an upstream functional node, they determine which downstream functional node the task should be routed to.

[0089] Workflow consists of nodes and edges, which are stored in computing devices using different data structures.

[0090] As an example, the data structure of a node is shown in Table 1 below:

[0091] Table 1

[0092]

[0093] As an example, the data structure of edges in the workflow is shown in Table 2 below:

[0094] Table 2

[0095]

[0096] S402. Obtain the modification suggestions input by the user in the first stage of the dialogue.

[0097] The modification suggestions describe directions for improvement for the initial workflow.

[0098] In this embodiment, the computing device can employ an information-gathering agent to engage in dialogue with the user to obtain modification suggestions input by the user. Specifically, the computing device can display an interactive interface on a display device, and the information-gathering agent can output dialogue on the interactive interface to guide the user to input modification suggestions.

[0099] In one possible implementation, S402 above can be specifically implemented as follows: obtaining the overall goal and modification suggestions of the target application; if the modification suggestions do not conform to the overall goal, prompting the user to re-enter the modification suggestions.

[0100] The overall goal refers to the main functions that the target application can achieve. When the information gathering agent receives modification suggestions from the user, it also determines whether the modification suggestions are consistent with the overall goal. If they are inconsistent, the user is prompted to re-enter the information. If they are consistent, the following step S403 is executed.

[0101] For example, if the overall goal of the target application is to achieve language translation, but the user's suggested modification is to enrich the colors of an image, it's clear that this requirement is completely irrelevant to the overall goal of language translation. Therefore, when the information-gathering agent receives the modification suggestion and determines that it does not conform to the overall goal, it prompts the user to re-enter the modification suggestion.

[0102] S403. The first intelligent agent is used to adjust the initial workflow to obtain candidate workflows.

[0103] The candidate workflow includes the target node; the target node is either a node modified based on the modification suggestions in the initial workflow, or a node added based on the modification suggestions in the initial workflow.

[0104] The first agent here can also be called the workflow adjustment agent.

[0105] It should be understood that dynamic workflow adjustment is a common requirement in software development and business process management. Traditional workflow adjustments require a significant amount of manual operation and expert intervention, which is not only time-consuming and error-prone but also demands a certain level of expertise from users. Therefore, this application embodiment employs an intelligent agent to adjust the workflow, making it easier for users to adjust the workflow, improving adjustment efficiency, and reducing errors.

[0106] In one possible implementation, prior to S403, the method provided in this application embodiment further includes: step a, using a first intelligent agent to generate a modification scheme based on modification suggestions, and displaying the modification scheme through a display device; step b, if the user rejects the modification scheme, the first intelligent agent regenerates the modification scheme.

[0107] When performing steps a-b above, S403 can be specifically implemented as follows: when the user confirms the modification plan, the first intelligent agent adjusts the initial workflow to the target workflow based on the modification plan.

[0108] In other words, when a user suggests modifications, the first intelligent agent can provide a preliminary modification plan based on the suggestion and display it to the user via a display device. If the user is not satisfied with the current modification plan, the first intelligent agent can regenerate the modification plan. If the user confirms the current modification plan, the first intelligent agent adjusts the initial workflow according to the current modification plan.

[0109] For example, during the development of translation software, a user discovers that the initial workflow cannot translate specialized terminology, and therefore suggests improvements (it cannot reflect specialized terminology; optimize it). Furthermore, the first agent, based on these suggestions, outputs a modification plan (adding a specialized terminology knowledge base to improve the accuracy of specialized data).

[0110] S404. The target workflow is obtained by using the second intelligent agent to configure the target node.

[0111] The target workflow conforms to the modification suggestions. The second agent can also be called the node configuration agent.

[0112] As mentioned earlier, the workflow used to build the application includes two levels of attributes for each node: a functional description and a functional implementation scheme. In step S403 above, the target node in the candidate workflow adjusted by the first agent based on the modification suggestions is newly added. It only describes the node's function at the functional description level; that is, the target node only contains the functional description attributes but not the specific functional implementation scheme. Therefore, the application cannot be built solely based on the candidate workflow. Therefore, this embodiment also executes step S404 above, configuring the functional implementation scheme for the target node through the second agent.

[0113] In some implementations, S404 can be specifically implemented as follows: a second agent determines a node template matching the modification suggestion from the node template set, and configures the functional implementation scheme of the target node using the node template. The node template set includes multiple node templates, and each node template includes a functional implementation scheme for one function.

[0114] It should be noted that after processing by the first agent, the functional description, inputs, and outputs of the target node have been defined (related to the modification suggestions). The node template set contains multiple preset node templates, each with corresponding functional description information. For the node configuration agent, its task is to match the functional description of each node in the initial workflow with the existing templates in the set; essentially, this is a clustering task. That is, for the target node, the node configuration agent performs semantic classification and recall based on the target node's description and the functional description information corresponding to each template, identifies semantically similar node templates, and configures the functional implementation scheme of those templates onto the target node.

[0115] Optionally, it is also necessary to consider whether the functional implementation scheme of the target node after searching and configuration can run normally. Therefore, in this embodiment, an additional intelligent agent (such as a functional verification intelligent agent) can be used to verify the target node.

[0116] For example, as described above, the target node has defined inputs and outputs, as well as corresponding functional implementation schemes. The functional verification agent can generate test information based on the target node's inputs and outputs, input the test information into the target node, and check whether the output meets expectations. If it does, it means that the functional implementation scheme of the target node can operate normally. If the output does not meet expectations, or if an error occurs, the functional verification agent can instruct the node configuration agent to re-search for node templates, or inform the user of the error so that the user can assist in making modifications.

[0117] In some implementations, if no node template matching the modification suggestion exists in the node template set, the computing device also performs the following: generating a function implementation scheme based on the function description of the target node using a third intelligent agent, and configuring the target node using the function implementation scheme.

[0118] The third agent can also be called the configuration-generating agent.

[0119] In other words, if there is no matching node in the existing set of node templates, a third agent can be used to generate a functional implementation plan based on the requirements (functional description) of the target node, ensuring that the target node can run.

[0120] S405. Generate the target application based on the target workflow.

[0121] In this embodiment, the computing device may employ a fifth intelligent agent to generate a target application based on a target workflow. This fifth intelligent agent is also referred to as an application-generating intelligent agent.

[0122] In one possible implementation, the application generation agent prompts the user to input the attribute information required to generate the application. This attribute information includes at least one of the following: name, version number, icon, and online status. Then, after obtaining the target application's attribute information, the computing device uses the application generation agent to generate the target application based on the target workflow and attribute information, completing the application building process. Additionally, the application generation agent can also generate an application identifier for the target application according to preset rules. This application identifier can be used to uniquely identify the application after it is launched on the application platform.

[0123] Optionally, the application-generating agent stores the application's attribute information in the application information table of the database to facilitate unified management of applications in the application platform.

[0124] Optionally, when the user inputs attribute information, the application generates an intelligent agent to determine whether the attribute information is missing. If it is missing, the user is prompted to supplement it.

[0125] In one possible implementation, the application construction method provided in this application embodiment further includes: using a fourth intelligent agent to verify the target workflow and determine whether the target workflow is running normally. If it is determined that the target workflow is running normally, a target application is generated based on the target workflow.

[0126] The fourth agent is also known as the workflow verification agent.

[0127] In some implementations, the above-mentioned use of a fourth intelligent agent to verify the target workflow includes: step 1, obtaining the initial input information of the target workflow; step 2, using the fourth intelligent agent to input the initial input information into the target workflow; and the fourth intelligent agent debugging each node sequentially according to the direction of the target workflow.

[0128] In other words, the process debugging agent can start from the beginning of the target workflow, input initial information, and when the user clicks "Next" on the front-end display device, the process debugging agent can execute the target workflow from the current node to the next node, output the debugging result of the next node, and display it to the user through the display device. If the debugging result of each node can be output and meets expectations (e.g., not empty value, or output error), it means that the target workflow can run normally.

[0129] In one possible implementation, when the target workflow fails to operate normally, a fault problem determined by a fourth agent is obtained; the fault problem is used to indicate the fault node and the cause of the fault; the first agent adjusts the target workflow based on the fault problem, and the fourth agent verifies the adjusted workflow until the target workflow operates normally.

[0130] In other words, if the target workflow fails to function correctly after being debugged by the fourth agent, the fourth agent can identify the fault during the debugging process. Then, the fourth agent can cooperate with the first agent, which adjusts the target workflow based on the fault, and the verification process is repeated. These steps are repeated until the target workflow functions correctly.

[0131] Figure 6 This is a schematic diagram of a workflow generation method provided in an embodiment of this application. Figure 6 As can be seen, there are two ways to generate workflows. One is to use a template matching agent to search for and determine a matching workflow template from a set of workflow templates. The other is to use a workflow generation agent to generate the workflow, which requires determining node templates from a set of node templates to configure the functional implementation details of the nodes in the generated workflow. Each node template in the set of node templates can be one or more combinations of prompts, large models, codes, RAGs, and tools.

[0132] It should be noted that different intelligent agents are used in the application construction process of this application embodiment, and different intelligent agents play a role in different stages of application construction. Each intelligent agent is obtained by training a large model using specific prompt words, and is essentially a workflow composed of some nodes and edges. As an example, the intelligent agent has the data structure shown in Table 3 below.

[0133] Table 3

[0134]

[0135] Figure 7 This is a schematic diagram illustrating a workflow adjustment process provided in an embodiment of this application. Figure 7 As shown, the initial workflow (including nodes 1 and 2), modification suggestions, and overall goals are input into the workflow adjustment module to obtain a new workflow (including nodes 1, 2, and 3). After the new workflow is validated by the workflow verification module, if any issues are found, the workflow adjustment module needs to make further adjustments. Finally, after verification confirms there are no problems, the modified target workflow is output.

[0136] Specifically, the workflow adjustment module includes an information gathering agent and a workflow adjustment agent. The information gathering agent interacts with the user to obtain the initial workflow, modification suggestions, and overall goals, and sends these to the workflow adjustment agent. The workflow adjustment agent, based on the node module set (search space) and knowledge base, adjusts the initial workflow to obtain the target workflow, which is then sent to the workflow verification module for further verification. The workflow verification module includes a workflow verification agent that can simulate workflow execution to verify the modified target workflow. If the target workflow verification identifies a problem, the workflow verification agent generates a fault issue and returns it, along with the failed workflows, to the workflow adjustment module for readjustment. This process iterates until the target workflow passes the verification of the workflow verification module, ensuring that the final target workflow meets user requirements and runs stably.

[0137] For example, Figure 8 This is a schematic diagram illustrating a specific example provided in an embodiment of this application. (In conjunction with...) Figure 8 Here's an explanation. If a user wants to develop a Chinese-to-English translation application, the initial workflow generated by the computing device is: acquire the input Chinese text - check if the input is compliant - acquire internet translation results / acquire large model translation results - refine the translation results - output the translation results. The user's suggestion is "It cannot translate technical terms; optimize it." Based on the user's suggestion, the workflow adjustment agent outputs the modification plan: "Add a technical terminology knowledge base to improve the accuracy of technical terminology translation." If the user confirms the modification plan, the workflow adjustment agent adjusts to obtain the target workflow: acquire the input Chinese text - check if the input is compliant - acquire internet translation results / acquire large model translation results / search for technical terminology knowledge base translations - refine the translation results - output the translation results.

[0138] In the application building method provided in this application embodiment, the computing device obtains an initial workflow and acquires modification suggestions from the user through a first-stage dialogue. Further, a first intelligent agent is used to process the initial workflow to obtain candidate workflows. Further, a second intelligent agent is used to configure the functional implementation scheme of the target nodes in the candidate workflows to obtain a target workflow that meets the modification suggestions, and then a target application is generated based on the target workflow. It can be seen that during application development and modification, users only need to input modification opinions, making the building process simple. The solution of this application can adjust the workflow flow according to the user's modification suggestions through the first intelligent agent and configure the functional implementation scheme of the workflow through the second intelligent agent, ultimately obtaining a target workflow that meets user needs and can run. Through the above method, the efficiency of application modification is effectively improved, and the complexity and possibility of errors in manual modification operations by users are reduced.

[0139] Furthermore, this solution utilizes intelligent agents to guide users in providing information related to workflow modifications through natural language interaction, thus addressing the issue of insufficient user guidance in existing technologies. Moreover, workflow modifications are collaboratively completed by multiple intelligent agents, each responsible for a specific task. Users only need to input their desired improvements (modification suggestions), requiring no further action, thus resolving the relatively complex nature of application modification.

[0140] The application construction method of this application embodiment can be widely applied to the following scenarios:

[0141] 1. Big Data Processing: Big data processing requires the construction of complex workflows to process and analyze data. The solution in this application embodiment can automatically adjust the workflow, improve efficiency, and reduce errors.

[0142] 2. Machine Learning and Artificial Intelligence: In the model training process of machine learning and artificial intelligence, it is often necessary to adjust the training process according to model performance and requirements. The solution in this application embodiment can automatically adjust these processes, thereby improving training efficiency.

[0143] 3. Software Development: During software development, developers need to adjust the development process according to project requirements. The solution in this application embodiment can automatically adjust the development process according to requirements, thereby improving development efficiency.

[0144] 4. Cloud Computing: In a cloud computing environment, service deployment and management often involve complex workflows. The solution in this application embodiment can automatically optimize these processes, improving the efficiency of cloud service deployment and management.

[0145] 5. Internet of Things (IoT): In an IoT environment, device management and data processing also require an effective workflow. The solutions in this application can help automatically optimize these processes and improve the operational efficiency of IoT systems.

[0146] As can be seen, the above mainly describes the solutions provided by the embodiments of this application from a methodological perspective. To achieve the above functions, the embodiments of this application provide corresponding hardware structures and / or software modules for executing each function. Those skilled in the art should readily recognize that, in conjunction with the modules and algorithm steps of the various examples described in the embodiments disclosed herein, the embodiments of this application can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed in hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0147] In an exemplary embodiment, this application also provides an application building apparatus. This application building apparatus may be the aforementioned computing device or a processor within the computing device. The application building apparatus may include one or more functional modules for implementing the application building method of the above method embodiments.

[0148] For example, Figure 9 This is a schematic diagram of an application building apparatus provided in an embodiment of this application. Figure 9 As shown, the application building apparatus includes an acquisition module 901 and a determination module 902. The acquisition module 901 and the determination module 902 are interconnected.

[0149] The acquisition module 901 is used to acquire an initial workflow in response to a modification request for the target application; the initial workflow is used to represent the workflow of the target application.

[0150] The acquisition module 901 is further configured to acquire modification suggestions input by the user during the first-stage dialogue; the modification suggestions are used to describe the direction of improvement for the initial workflow;

[0151] The determining module 902 is used to adjust the initial workflow using a first intelligent agent to obtain a candidate workflow; the candidate workflow includes a target node; the target node is a node in the initial workflow that has been modified based on the modification suggestion, or a node in the initial workflow that has been added based on the modification suggestion;

[0152] The determining module 902 is further configured to use a second intelligent agent to configure the functional implementation scheme of the target node to obtain the target workflow; the target workflow conforms to the modification suggestion;

[0153] The determining module 902 is further configured to generate the target application based on the target workflow.

[0154] In one possible implementation, the determining module 902 is specifically used to: use the second intelligent agent to determine a node template that matches the modification suggestion from the node template set, and use the node template to configure the functional implementation scheme of the target node; the node template set includes multiple node templates, and each node template includes a functional implementation scheme of a function.

[0155] In another possible implementation, if no node template matching the modification suggestion exists in the node template set, the determining module 902 is further configured to use a third agent to generate a function implementation scheme based on the function description of the target node, and configure the target node using the function implementation scheme.

[0156] In another possible implementation, the determining module 902 is further configured to use a fourth intelligent agent to verify the target workflow and determine whether the target workflow is running normally; specifically, the determining module 902 is configured to generate the target application based on the target workflow if the target workflow is determined to be running normally.

[0157] In another possible implementation, the determining module 902 is specifically used to obtain the initial input information of the target workflow;

[0158] The fourth intelligent agent inputs the initial input information into the target workflow; the fourth intelligent agent debugs each node sequentially according to the direction of the target workflow.

[0159] In another possible implementation, the acquisition module 901 is further configured to acquire a fault problem determined by the fourth agent when the target workflow cannot operate normally; the fault problem is used to indicate the fault node and the cause of the fault; the determination module 902 is further configured to use the first agent to adjust the target workflow based on the fault problem, and use the fourth agent to verify the adjusted workflow until the target workflow operates normally.

[0160] In another possible implementation, the determining module 902 is further configured to: use a first intelligent agent to generate a modification scheme based on the modification suggestion, and display the modification scheme through a display device; if the user rejects the modification scheme, the first intelligent agent regenerates the modification scheme; specifically, the determining module 902 is configured to: if the user confirms the modification scheme, use the first intelligent agent to adjust the initial workflow to the target workflow based on the modification scheme.

[0161] In another possible implementation, the acquisition module 901 is specifically used to acquire the overall goal of the target application and the modification suggestions; if the modification suggestions do not conform to the overall goal, the user is prompted to re-enter the modification suggestions.

[0162] In another possible implementation, the determining module 902 is specifically used to obtain the attribute information of the target application; the attribute information includes at least one of the following: name, version number, icon, online status; and a fifth intelligent agent generates the target application based on the target workflow and the attribute information.

[0163] In an exemplary embodiment, this application also provides a computing device. Figure 10 This is a schematic diagram illustrating the composition of a computing device provided in an embodiment of this application. Figure 10 As shown, the computing device may include a processor 1001 and a memory 1002; the memory 1002 stores instructions executable by the processor 1001; when the processor 1001 is configured to execute instructions, the computing device implements the method described in the foregoing method embodiments.

[0164] This application also provides a computer-readable storage medium. All or part of the processes in the above method embodiments can be executed by computer instructions instructing related hardware; for example, the related hardware can be a processor of a computing device. The program instructions can be stored in the above-described computer-readable storage medium, and when executed, they can implement the processes of the above method embodiments. The computer-readable storage medium can be memory. The above-described computer-readable storage medium can also be an external storage device, such as a hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc. Further, the above-described computer-readable storage medium can include both memory and external storage devices. The above-described computer-readable storage medium is used to store the above-described computer program instructions and other programs and data required for the above-described software package translation.

[0165] This application also provides a computer program product comprising a computer program that, when run on a computing device, causes the computing device to execute any of the software package translation methods provided in the above embodiments.

[0166] Although this application has been described herein in conjunction with various embodiments, those skilled in the art, by reviewing the accompanying drawings, disclosure, and appended claims, will understand and implement other variations of the disclosed embodiments in carrying out the claimed application. In the claims, the word "comprising" does not exclude other components or steps, and "a" or "an" does not exclude multiple components. A single processor or other unit can implement several functions listed in the claims. While different dependent claims may recite certain measures, this does not mean that these measures cannot be combined to produce good results.

[0167] Although this application has been described in conjunction with specific features and embodiments, it is obvious that various modifications and combinations can be made thereto without departing from the spirit and scope of this application. Accordingly, this specification and drawings are merely exemplary illustrations of this application as defined by the appended claims, and are considered to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from the spirit and scope of this application. Thus, if such modifications and modifications of this application fall within the scope of the claims of this application and their equivalents, this application is also intended to include such modifications and modifications.

[0168] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. An application construction method, characterized in that, The method includes: In response to modification requests for the target application, an initial workflow is obtained; the initial workflow represents the workflow of the target application. Obtain modification suggestions input by the user during the first phase of the dialogue; the modification suggestions are used to describe directions for improvement for the initial workflow; The initial workflow is adjusted using a first intelligent agent to obtain a candidate workflow; the candidate workflow includes a target node; the target node is a node in the initial workflow that has been modified based on the modification suggestion, or a node in the initial workflow that has been added based on the modification suggestion. The target workflow is obtained by configuring the function implementation scheme of the target node using a second intelligent agent; the target workflow conforms to the modification suggestion. Based on the target workflow, the target application is generated.

2. The method according to claim 1, characterized in that, The implementation scheme of configuring the target node using a second intelligent agent yields the target workflow, including: The second intelligent agent determines a node template that matches the modification suggestion from the node template set, and uses the node template to configure the functional implementation scheme of the target node; the node template set includes multiple node templates, and each node template includes a functional implementation scheme of one function.

3. The method according to claim 2, characterized in that, If no node template matching the modification suggestion exists in the set of node templates, the method further includes: A third intelligent agent generates a function implementation scheme based on the function description of the target node, and the target node is configured using the function implementation scheme.

4. The method according to any one of claims 1-3, characterized in that, The method further includes: A fourth intelligent agent is used to verify the target workflow and determine whether the target workflow is running normally; Based on the target workflow, the target application is generated, including: If the target workflow is confirmed to be operating normally, the target application is generated based on the target workflow.

5. The method according to claim 4, characterized in that, The step of using a fourth intelligent agent to verify the target workflow includes: Obtain the initial input information of the target workflow; The fourth intelligent agent inputs the initial input information into the target workflow; the fourth intelligent agent debugs each node sequentially according to the direction of the target workflow.

6. The method according to claim 4 or 5, characterized in that, The method further includes: If the target workflow fails to operate normally, the fault problem determined by the fourth agent is obtained; the fault problem is used to indicate the fault node and the cause of the fault. The first intelligent agent adjusts the target workflow based on the fault problem, and the fourth intelligent agent verifies the adjusted workflow until the target workflow is running normally.

7. The method according to any one of claims 1-6, characterized in that, The method further includes: A first intelligent agent generates a modification scheme based on the modification suggestions, and displays the modification scheme through a display device; If the user rejects the proposed modification, the first agent will regenerate the modification. Employing a first intelligent agent to adjust the initial workflow into the target workflow includes: If the user confirms the modification plan, the first intelligent agent adjusts the initial workflow to the target workflow based on the modification plan.

8. The method according to claim 7, characterized in that, The process of obtaining modification suggestions input by the user in the first stage of the dialogue includes: Obtain the overall goals of the target application and the proposed modifications; If the proposed modifications do not meet the overall objective, the user will be prompted to re-enter the proposed modifications.

9. The method according to claim 1, characterized in that, Based on the target workflow, the target application is generated, including: Obtain the attribute information of the target application; the attribute information includes at least one of the following: name, version number, icon, online status; The fifth intelligent agent generates the target application based on the target workflow and the attribute information.

10. A computing device, characterized in that, The computing device includes a processor and a memory; the processor is coupled to the memory. The memory is used to store computer instructions; The computer instructions are loaded and executed by the processor to enable the computing device to implement the application building method as described in any one of claims 1-9.

Images