An AI-driven multi-industry DevOps full-process intelligent method, device and medium

Abstract

The application discloses an AI-driven multi-industry DevOps full-process intelligent method, device and medium, which comprises the following steps: connecting a DevOps tool chain to extract full-process key data and construct a structured data asset pool; building a layered architecture combining a basic model and an extended model; built-in multi-industry characteristic plug-ins are deployed and dynamically updated according to business scenarios; the plug-ins interact with AI models in real time to output results conforming to industry characteristics; user operation logs are monitored, development scenarios are identified, knowledge assets are retrieved from the data asset pool, and contents are actively pushed with source identification; potential risks are analyzed in combination with associated data, a risk correlation model is established to predict the subsequent influence range, an early warning is generated and pushed to relevant roles, and the risk processing state is tracked to form a closed-loop management. Through data integration, layered AI architecture, active knowledge empowerment and risk closed-loop management, the application realizes DevOps full-process intelligence and multi-industry adaptation, and improves the development efficiency and process stability.

Description

TECHNICAL FIELD

[0001] Embodiments of the present application relate to the field of DevOps technology and AI technology, in particular to a method and device for AI-driven intelligentization of multi-industry DevOps full process, and a medium. BACKGROUND

[0002] With the wide application of the DevOps concept in various industries, the DevOps tool chain introduced in the enterprise R&D process is increasingly complex, covering multiple links such as demand collaboration, code management, building and deployment, testing and operation, etc. However, the related technology has the following key problems:

[0003] 1. The data of each DevOps tool is isolated, and there is a lack of unified association identifier across links, resulting in difficulty in integrating full-process data and forming effective data assets;

[0004] 2. The existing AI capabilities in the DevOps scenario are mostly single-task oriented, lack hierarchical architecture design, and are difficult to adapt to the diversified needs of different industries;

[0005] 3. There is a lack of active knowledge pushing mechanism in the R&D process, and the knowledge is scattered in various knowledge bases, so users need to manually search for related information, which is inefficient;

[0006] 4. Risk identification and processing are mostly limited to a single link, cannot predict the impact of risks on subsequent links, and lack closed-loop management, with a high rate of repeated occurrence of similar risks.

[0007] In view of the above problems, no effective solution has been proposed so far. Therefore, there is an urgent need for a technical solution that can realize intelligentization of DevOps full process, adapt to the needs of multiple industries, and have risk closed-loop management capability. SUMMARY

[0008] Embodiments of the present application provide a method, device and medium for AI-driven intelligentization of multi-industry DevOps full process, to solve the problems of related technology. The technical solution is as follows:

[0009] In a first aspect, the embodiments of the present application provide a method for AI-driven intelligentization of multi-industry DevOps full process, comprising:

[0010] Integrate the DevOps tool chain, extract key data of the full process, establish a unique identifier across links and associate the key data, and build a structured data asset pool after processing the extracted data;

[0011] Build a hierarchical architecture of AI model combining a basic model and an extended model, deploy standardized AI general capabilities in the basic model layer, and provide a plug-in interface in the extended model layer to automatically achieve different intelligent analysis effects according to different AI task types;

[0012] Built-in industry characteristic plug-ins, one-key deployment and dynamic update plug-ins according to business scenarios, so that the plug-ins can interact with AI models in real time and automatically load corresponding industry rules when processing R&D tasks, and output results conforming to industry characteristics;

[0013] Monitoring user operation logs in DevOps tools, identifying R&D scenarios and retrieving matching knowledge assets from the structured data asset pool, actively pushing content through tool link interfaces in the user operation interface and attaching source identification to support one-key tracing to the original document;

[0014] Combining unique identification associated key data, analyzing potential risks and establishing a risk correlation model, automatically predicting the impact range on subsequent links when detecting risks in a certain link, generating early warning information and processing suggestions and pushing them to relevant roles, while tracking risk processing status to form a risk closed-loop management, reducing the recurrence rate of similar risks.

[0015] In one implementation method, the structured data asset pool is constructed by:

[0016] Connecting the DevOps tool chain, extracting key data throughout the process through API interfaces or data collection plug-ins;

[0017] Establishing a unique identifier across links and associating the unique identifier with the extracted key data to achieve cross-link data correlation;

[0018] After cleaning, converting, and integrating the extracted raw data, a structured data asset pool is constructed to provide a data foundation for subsequent AI analysis and knowledge retrieval.

[0019] In one implementation method, the AI model hierarchical architecture is constructed by:

[0020] The basic model layer deploys standardized AI general capabilities, including natural language processing models, machine learning models, and knowledge graph models; the basic model layer is trained with large-scale general data and has cross-scenario basic intelligent analysis capabilities;

[0021] The extended model layer provides a standardized plug-in interface to support the connection of customized model plug-ins according to different AI task types, and automatically realizes different intelligent analysis effects.

[0022] In one implementation method, the built-in industry characteristic plug-in that outputs results conforming to industry characteristics includes:

[0023] The built-in multi-industry characteristic plug-in contains business rule libraries, scenario template libraries, and R&D process templates for each industry;

[0024] When using it, you can deploy the required industry plugins with one click through the visual interface according to the actual business scenario;

[0025] When industry rules or business needs change, the plugin can automatically update the rule base and functional modules through an online upgrade mechanism.

[0026] When handling R&D tasks, the plugin interacts with the AI ​​model in real time, automatically inputting industry rules into the AI ​​model as analysis constraints, so that the AI ​​model outputs results that conform to industry characteristics.

[0027] In one implementation method, the monitoring of user operation logs in DevOps tools, including proactively pushing content with a source identifier, includes:

[0028] Real-time monitoring of user operation behavior logs in the DevOps toolchain, and identification of corresponding development scenarios through behavior analysis algorithms;

[0029] Based on the identified R&D scenario, retrieve matching knowledge assets from the structured data asset pool;

[0030] Through the DevOps toolchain's interface, relevant knowledge asset content is proactively pushed to the user interface, along with source identification.

[0031] When users click on the pushed content, they can trace back to the original document with one click through the source identifier, quickly obtain complete information, and improve the efficiency of knowledge acquisition in the research and development process.

[0032] In one implementation method, the risk closed-loop management includes:

[0033] Potential risks are analyzed using risk identification algorithms, and risk correlation models are established.

[0034] The risk association model, based on the correlation between historical risk data and full-process data, can trace the associated data of a risk by using a unique identifier across stages when a risk is detected, and automatically predict the scope of the risk's impact on subsequent stages.

[0035] The system generates early warning information based on the risk level and scope of impact, and pushes it to the work interface or message terminal of the relevant roles;

[0036] The system tracks the status of risk handling and records the process and results. It then sends the risk handling process and results data back to the structured data asset pool to optimize the risk association model and further reduce the recurrence rate of similar risks.

[0037] Secondly, embodiments of this application also provide an AI-driven device for intelligentizing the entire DevOps process across multiple industries, comprising:

[0038] A data asset construction module is configured to connect to a DevOps tool chain, extract key data in the whole process, establish a unique identifier across links and associate the key data, and construct a structured data asset pool after processing the extracted data;

[0039] An AI model architecture module is configured to construct an AI model layered architecture combining a basic model and an extended model, the basic model layer deploys standardized AI general capabilities, and the extended model layer provides a plug-in interface to automatically achieve different intelligent analysis effects according to different AI task types;

[0040] An industry plug-in management module is configured to internally store industry characteristic plug-ins, one-key deploy and dynamically update the plug-ins according to business scenarios, so that the plug-ins can interact with AI models in real time and automatically load corresponding industry rules when processing R&D tasks, and output results conforming to industry characteristics;

[0041] A knowledge pushing module is configured to monitor user operation logs in DevOps tools, identify R&D scenarios, and retrieve matching knowledge assets from a structured data asset pool, actively push content to a user operation interface through a tool link and attach a source identifier, to support one-key tracing to original documents;

[0042] A risk closed-loop management module is configured to analyze potential risks and establish a risk correlation model based on the unique identifier associated key data, automatically predict the impact range on subsequent links when detecting risks in a certain link, generate early warning information and processing suggestions and push them to relevant roles, and track processing status to form closed-loop management and reduce the recurrence rate of similar risks.

[0043] In a third aspect, the embodiments of the present application also provide a computer readable storage medium, the computer readable storage medium stores a computer program, and the computer program is executed by a processor to implement the method in any one of the embodiments of the above aspects.

[0044] The advantages or beneficial effects of the above technical solutions at least include:

[0045] Break the data island of DevOps tools and form high-value data assets: by establishing a unique identifier (such as "project ID-task ID-time stamp") across links to associate key data in the whole process, and combining data cleaning, conversion, integration processing to construct a structured data asset pool, the problem of data isolation in traditional DevOps tools and the inability to form effective assets is solved, providing a unified data foundation for subsequent AI analysis and knowledge retrieval, and improving the data reuse value;

[0046] Implement AI capability hierarchical adaptation to meet diversified needs of multiple industries: build a hierarchical architecture of "basic model layer + extended model layer", the basic model layer provides standardized general AI capabilities (natural language processing, machine learning, knowledge graph), and the extended model layer supports customized model access through plug-in interface, which not only guarantees cross-scenario basic intelligent capabilities, but also adapts to different AI tasks such as code defect detection and build failure root cause analysis, solving the problem of single task orientation and poor adaptability of traditional AI in DevOps scenarios;

[0047] Dynamically adapt to industry characteristics to reduce industry landing costs: built-in industry-specific plug-ins (including business rule library, scenario template library, and R&D process template) covering finance, healthcare, manufacturing, and Internet, support one-key deployment and online dynamic update, plug-ins and AI models interactively load industry rules in real time, without the need to restructure the underlying system for different industries, significantly reducing the development cost and cycle of DevOps in multiple industries;

[0048] Improve R&D knowledge acquisition efficiency and reduce labor costs: identify R&D scenarios by monitoring user operation logs, actively push matching knowledge assets with source identification, support one-key tracing to original documents, solve the problem of scattered knowledge and manual retrieval in traditional R&D, shorten the knowledge acquisition time of R&D personnel, and improve R&D efficiency;

[0049] Implement risk full-process closed-loop management to reduce the recurrence rate of similar risks: build a risk association model based on unique identification of associated data, predict the impact of risks on subsequent links, generate early warning and processing suggestions and track processing status, form a "identification-warning-processing-recording-optimization" closed loop, solve the problem of traditional risk management limited to a single link without closed loop, and significantly reduce the recurrence rate of similar risks.

[0050] The above summary is only for the purpose of the description and is not intended to limit in any way. In addition to the illustrative aspects, embodiments, and features described above, further aspects, embodiments, and features will be readily apparent to those skilled in the art by reference to the drawings and the following detailed description. BRIEF DESCRIPTION OF DRAWINGS

[0051] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed in the embodiments will be briefly introduced below. In the drawings, the same reference signs represent the same or similar parts or elements throughout the drawings unless otherwise specified. These drawings are not necessarily drawn to scale. It should be understood that the following drawings only show some embodiments of the present application, and therefore should not be considered as limiting the scope, and other related drawings can be obtained by those skilled in the art without creative labor.

[0052] Figure 1 A flowchart of a method for AI-driven multi-industry DevOps full-process intelligentization according to an embodiment of the present application is provided.

[0053] Figure 2 A structured data asset pool construction flowchart according to an embodiment of the present application is provided.

[0054] Figure 3 An AI model hierarchical architecture diagram according to an embodiment of the present application is provided.

[0055] Figure 4 A one-key deployment and dynamic update plug-in flowchart according to a business scenario according to an embodiment of the present application is provided.

[0056] Figure 5 An active knowledge push and traceability flowchart according to an embodiment of the present application is provided.

[0057] Figure 6 A risk closed-loop management flowchart according to an embodiment of the present application is provided.

[0058] Figure 7 A pipeline AI risk early warning and suggestion diagram according to an embodiment of the present application is provided.

[0059] Figure 8 An industry demand generation effect diagram according to an embodiment of the present application is provided.

[0060] Figure 9 An industry test case generation effect diagram according to an embodiment of the present application is provided.

[0061] Figure 10 A code inspection AI risk early warning and suggestion diagram according to an embodiment of the present application is provided.

[0062] Figure 11 A device module structure diagram according to an embodiment of the present application is provided. DETAILED DESCRIPTION

[0063] In the following, only certain exemplary embodiments are simply described. As those skilled in the art can recognize, the described embodiments can be modified in various different ways without departing from the spirit or scope of the present application. Therefore, the drawings and the description are considered to be exemplary in nature rather than limiting.

[0064] The present application discloses a method for AI-driven multi-industry DevOps full-process intelligentization, the method flowchart is as shown in Figure 1 The following steps are included:

[0065] S101, connect to the DevOps tool chain, extract key data throughout the process, establish a unique identifier across the link and associate the key data, and build a structured data asset pool after processing the extracted data.

[0066] In specific implementation, the DevOps tool chain is connected, including the collaborative management tool CTeam, the code management tool CCode, the compilation and construction tool CCI, the test tool CTest, the artifact management tool CPack, and the operation and monitoring tool. Through API interface or data collection plug-in, key data throughout the process is extracted, including code submission record (including submitter, submission time, code change content), build log (including build steps, dependent package information, build result), test report (including test case execution result, defect level and description), and operation index data (including server CPU usage, memory occupancy, transaction response time). To realize cross-link data association, a cross-link unique identifier based on "project ID-task ID-time stamp" is generated, and the unique identifier is associated and bound with the extracted key data. After cleaning (removing redundant data, correcting abnormal values), conversion (unifying data format and field naming), and integration (associating data of different links according to business logic), the original data is processed to build a structured data asset pool as shown in Figure 2 , which provides a data basis for subsequent AI analysis and knowledge retrieval.

[0067] S102, build an AI model layered architecture combining basic models and extended models, the basic model layer deploys standardized AI general capabilities, and the extended model layer provides plug-in interfaces to automatically achieve different intelligent analysis effects according to different AI task types.

[0068] In specific implementation, the AI model layered architecture diagram is as shown in Figure 3 , the basic model layer deploys standardized AI general capabilities, including BERT natural language processing model (used for parsing requirement documents, test defect descriptions and operation log texts), LSTM machine learning model (used for index prediction and anomaly detection), Neo4j knowledge graph model (used for building "code module-defect type-solution" research and development knowledge association network), etc. This layer of model is trained with large-scale general data and has cross-scenario basic intelligent analysis capability; the extended model layer provides standardized plug-in interfaces to support the access of customized model plug-ins according to different AI task types (such as code defect detection, build failure root cause analysis, operation fault prediction, etc.), and automatically achieve different intelligent analysis effects.

[0069] S103, built-in industry characteristic plug-in, one-key deployment and dynamic update plug-in according to business scenarios, so that the plug-in can interact with the AI model in real time and automatically load the corresponding industry rules when processing research and development tasks, and output results conforming to industry characteristics.

[0070] The built-in industry characteristic plug-ins cover multiple industries such as finance, medical care, manufacturing, and the Internet, and contain the business rule library of each industry (such as the compliance verification rules of the financial industry and the privacy data protection rules of the medical industry), the scene template library (the code audit scene of the financial industry), and the R&D process template (such as the hardware R&D process template of the manufacturing industry and the agile development process template of the Internet industry), all of which are derived from the summary of the DevOps practice experience of industry clients served by the company for more than ten years. Users can deploy the required industry plug-ins through a visual interface according to the actual business scene, as shown in the process diagram Figure 5 Meanwhile, the system supports dynamic updating of plug-ins, and when the industry rules or business requirements change, the plug-ins can automatically update the rule library and functional modules through an online upgrade mechanism.

[0071] For example, the financial industry characteristic DevOps plug-in package is deployed for a financial client. The plug-in package contains a “compliance code detection plug-in” that can detect whether there are problems such as sensitive data not being encrypted and illegal transaction logic in the code based on the financial industry compliance rule library, as shown in Figure 7 The “operation and maintenance risk prediction plug-in” predicts the risk of server resource shortage based on the characteristics of the peak period of financial transactions.

[0072] When processing R&D tasks, the plug-in interacts with the AI model in real time, automatically loads the plug-in assembled in the extended model interface, inputs the plug-in content into the model, and realizes different intelligent analysis effects according to different task types.

[0073] For example, in the requirement generation task, the AI model can assist in understanding the requirement generation prompt input by the user, identify industry-specific terms in the prompt content, and match the corresponding industry-specific plug-in. Combined with the industry scene template library and the business rule library, it can generate a document that meets the industry functional requirements. The specific interface effect can be referred to Figure 8 In the test case generation task, the AI identifies the requirement content that needs to generate a test case, and combines the R&D process template in the industry characteristic plug-in to automatically generate a standardized test case. The specific interface effect can be referred to Figure 9 In the code development task, the “compliance code detection plug-in” interacts with the model in real time, inputs the compliance rules into the AI model as an analysis constraint condition, and when the model detects that a certain piece of code does not encrypt the customer’s ID number, it outputs a rectification suggestion that meets the financial compliance requirements, such as “suggest using the AES-256 encryption algorithm to encrypt the ID number field”. The specific interface effect can be referred to Figure 10 .

[0074] S104, monitor user operation logs in DevOps tools, identify development scenarios and retrieve matching knowledge assets from structured data asset pools, actively push content through tool link interfaces in user operation interfaces and attach source identification to support one-click tracing to original documents.

[0075] Real-time monitoring of user operation logs in DevOps tools (such as user code submission in CCode, build triggering in CCI, test case execution in CTest, etc.), identifying current development scenarios (such as code development scenario, build deployment scenario, test verification scenario) through behavior analysis algorithm. According to the identified development scenario, retrieve matching knowledge assets from structured data asset pool, and through the interface of DevOps tool chain, actively push related knowledge asset content in user operation interface, while attaching source identification, including original document ID, creation time, author information, etc. Active knowledge pushing and tracing process as shown in Figure 4 When the user clicks on the pushed content, it can be traced back to the original document through the source identification, quickly obtain complete information, and improve the efficiency of knowledge acquisition in the development process.

[0076] For example, after monitoring that the developer submits the "customer account management module" code in CCode, the scenario recognition unit determines that the current scenario is code development and verification. The knowledge retrieval unit retrieves matching knowledge assets from the data asset pool and finds "historical customer account module code defect solution" and "financial industry account data encryption technology document". Through the CCode interface, the knowledge assets are actively pushed, with source identification "DOC-FIN-028 (20240315, R&D Center Technology Department)". The developer clicks on the pushed content to directly open the corresponding document address and quickly obtain the encryption implementation details.

[0077] S105, combined with the key data associated with the unique identification, analyze potential risks and establish risk correlation model, automatically predict the influence range on subsequent links when detecting certain link risk, generate warning information and processing suggestions and push to related roles, and track risk processing status to form closed-loop management, reduce the recurrence rate of similar risks.

[0078] The potential risks are analyzed by a risk identification algorithm, and a risk correlation model is established. The risk correlation model is based on the correlation between historical risk data and full-process data. When a risk in a certain link is detected, the influence range of the risk on subsequent links can be automatically predicted (for example, a code defect risk may affect the subsequent build success rate and test pass rate, and an insufficient operation and maintenance resource risk may affect the subsequent service deployment and user experience). The system generates early warning information according to the risk level and influence range, and pushes it to the work interface or message terminal of the relevant roles. At the same time, the system tracks the risk processing state, promotes the relevant roles to complete risk processing through a task coordination mechanism, and records the processing process and results, forming a risk closed-loop management as shown in FIG. 8, which reduces the recurrence rate of similar risks. Figure 6

[0079] For example, the risk identification module analyzes the operation and maintenance index data through an LSTM model, finds that the "customer transaction response time" has increased from the normal 0.5 seconds to 2 seconds, which exceeds the threshold of 1.5 seconds, and combines the unique identification associated with the build record to find that the anomaly occurs after the deployment of the latest version. The risk correlation model predicts that the risk will affect the "customer transaction success rate" in the subsequent process, which may cause transaction timeout failure. An early warning information is generated (risk description: "abnormal increase in transaction response time"; affected links: transaction processing, user experience; risk level: serious; recommended measures: rollback to the previous stable version, and troubleshoot the performance bottleneck in the new version code), which is pushed to the WeChat terminal of the operation and maintenance personnel and the project manager. After receiving the early warning, the operation and maintenance personnel perform version rollback operation, and update the processing state to "processing" in the system; after troubleshooting, it is found that the newly added log printing logic in the new version occupies too much resource, and after repair, it is redeployed, and the state is updated to "solved"; the project manager updates the state to "closed" after confirming that the risk is eliminated. The system records the entire processing process to form a risk closed loop, which is used to reduce the occurrence rate of similar performance risks caused by log printing.

[0080] Figure 11 An apparatus module structure diagram according to an embodiment of the present application is shown, which comprises:

[0081] A data asset construction module is configured to interface with a DevOps tool chain, extract full-process key data, establish a unique identification across links and associate the key data, and construct a structured data asset pool after processing the extracted data;

[0082] An AI model architecture module is configured to construct an AI model layered architecture combining a basic model and an extended model. The basic model layer deploys standardized AI general capabilities, and the extended model layer provides a plug-in interface to automatically achieve different intelligent analysis effects according to different AI task types.

[0083] ​An industry plug-in management module is configured to embed an industry characteristic plug-in, deploy and dynamically update the plug-in according to a business scenario, and enable the plug-in to interact with an AI model in real time and automatically load corresponding industry rules when processing a research and development task, so as to output a result conforming to an industry characteristic.

[0084] A knowledge pushing module is configured to monitor a user operation log in a DevOps tool, identify a research and development scenario, and retrieve a matched knowledge asset from a structured data asset pool, and actively push content to a user operation interface through a tool link and attach a source identifier, so as to support one-key tracing to an original document.

[0085] A risk closed loop management module is configured to combine unique identification associated key data, analyze potential risks, and establish a risk correlation model, automatically predict an influence range on a subsequent link when detecting a risk in a certain link, generate early warning information and processing suggestions, and push the early warning information and the processing suggestions to related roles, and track a processing state to form a closed loop management, and reduce a same risk repetition rate.

[0086] The embodiment of the present application provides a computer readable storage medium, and the computer readable storage medium stores a computer program. When the computer program runs on a computer, the method provided in the embodiment of the present application is implemented.

[0087] The embodiment of the present application further provides a chip, which includes a processor, and is configured to call and run instructions stored in a memory, so that a communication device installed with the chip executes the method provided in the embodiment of the present application.

[0088] The embodiment of the present application further provides a chip, which includes an input interface, an output interface, a processor and a memory, and the input interface, the output interface, the processor and the memory are connected through an internal connection path. The processor is configured to execute code in the memory. When the code is executed, the processor is configured to execute the method provided in the embodiment of the present application.

[0089] It is to be understood that the above-described processor can be a Central Processing Unit (CPU), and can also be other general-purpose processors, a Digital Signal Processing (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other programmable logic device, discrete gate or transistor logic device, discrete hardware component, and the like. The general-purpose processor can be a microprocessor or any conventional processor, and the like. It is to be noted that the processor can be a processor supporting an Advanced RISC Machines (ARM) architecture.

[0090] Further, the memory can include a read-only memory and a random access memory, and can further include a non-volatile random access memory. The memory can be a volatile memory or a non-volatile memory, or can include both volatile and non-volatile memories. The non-volatile memory can include a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory can include a random access memory (RAM) used as an external cache. By way of example, and not limitation, many forms of RAM can be used. For example, a static random access memory (SRAM), a dynamic random access memory (DRAM), a synchronous dynamic random access memory (SDRAM), a double data rate synchronous dynamic random access memory (DDR SDRAM), an enhanced SDRAM (ESDRAM), a synchlink DRAM (LDRAM), and a direct rambus RAM (DR RAM) can be used.

[0091] In the above-described embodiments, all or part can be implemented by software, hardware, firmware, or any combination thereof. When implemented by software, all or part can be implemented in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded on a computer, all or part generates a flow or function according to the present application. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transferred from one computer-readable storage medium to another computer-readable storage medium.

[0092] In the description of the present specification, the description of the terms "one embodiment", "some embodiments", "an example", "a specific example", or "some examples" and the like means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. Also, the specific features, structures, materials or characteristics described can be combined in any appropriate manner in any one or more embodiments or examples. In addition, the person skilled in the art can combine and combine the different embodiments or examples described in the present specification and the features of the different embodiments or examples without contradiction.

[0093] In addition, the terms "first", "second", etc. are used only for the purpose of description and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined with "first", "second" can explicitly or implicitly include at least one of the features. In the description of the present application, the meaning of "a plurality of" is two or more, unless otherwise explicitly specified.

[0094] Any process or method descriptions in flow charts or described elsewhere herein can be understood as representing code modules, segments, or portions of code that include one or more executable instructions for implementing specific logic functions or other processes. Also, the preferred embodiments of the present application can include additional implementation with respect to the order in which steps are performed, including simultaneous or substantially simultaneous performance of functions, or in reverse order of the steps shown or discussed.

[0095] The logic and / or steps represented in flow charts or otherwise described herein, for example, can be considered as a list of executable instructions for implementing logic functions, which can be specifically embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus or device, such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, apparatus or device and execute the instructions, or in conjunction with these instructions execution system, apparatus or device.

[0096] It should be understood that each part of the present application can be realized by hardware, software, firmware or a combination thereof. In the above embodiments, a plurality of steps or methods can be realized by software or firmware stored in a memory and executed by a suitable instruction execution system. All or part of the steps of the above-mentioned embodiment methods can be completed by a program instructing the relevant hardware, which can be stored in a computer readable storage medium and includes one or a combination of the steps of the embodiment methods when executed.

[0097] In addition, each functional unit in each embodiment of the present application can be integrated in one processing module, or each unit can be physically present separately, or two or more units can be integrated in one module. The above-mentioned integrated module can be realized in the form of hardware or in the form of a software functional module. The above-mentioned integrated module, if realized in the form of a software functional module and sold or used as an independent product, can also be stored in a computer readable storage medium. The storage medium can be a read-only memory, a magnetic disk or an optical disk, etc.

[0098] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of various changes or replacements within the technical scope disclosed in the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for AI-driven intelligentization of multi-industry DevOps full-process, characterized in that, The application comprises: connecting a DevOps tool chain, extracting key data throughout the process, establishing a unique identifier across the links and associating the key data, and constructing a structured data asset pool after processing the extracted data; constructing an AI model layered architecture combining basic models and extended models, deploying standardized AI general capabilities in the basic model layer, and providing plug-in interfaces in the extended model layer to automatically achieve different intelligent analysis effects according to different AI task types; built-in industry-specific plug-ins, one-click deployment and dynamic update of plug-ins according to business scenarios, enabling plug-ins to interact with AI models in real time and automatically load corresponding industry rules when processing R&D tasks, and output results that meet industry-specific characteristics; monitoring user operation logs in DevOps tools, identifying R&D scenarios and retrieving matching knowledge assets from the structured data asset pool, actively pushing content through tool link interfaces in the user operation interface and attaching source identifiers to support one-click tracing to the original document; combined with the key data associated with the unique identifier, analyze potential risks and establish a risk correlation model, automatically predict the impact range on subsequent links when detecting risks in a certain link, generate warning information and processing suggestions and push them to relevant roles, and track risk processing status to form a risk closed-loop management, reducing the recurrence rate of similar risks.

2. The method of claim 1, wherein, The construction of the structured data asset pool comprises: connecting a DevOps tool chain, extracting key data throughout the process through API interfaces or data collection plug-ins; establishing a unique identifier across the links, and associating the unique identifier with the extracted key data to achieve cross-link data association; after cleaning, converting, and integrating the extracted raw data, a structured data asset pool is constructed to provide a data foundation for subsequent AI analysis and knowledge retrieval.

3. The method of claim 1, wherein, The construction of the AI model layered architecture comprises: the basic model layer deploys standardized AI general capabilities, including natural language processing models, machine learning models, and knowledge graph models; the basic model layer is trained with large-scale general data and has cross-scenario basic intelligent analysis capabilities; the extended model layer provides standardized plug-in interfaces to support the connection of customized model plug-ins according to different AI task types, automatically achieving different intelligent analysis effects.

4. The method of claim 1, wherein, The built-in industry-specific plug-ins output results that meet industry-specific characteristics, including: built-in multi-industry plug-ins contain industry-specific rule libraries, scenario template libraries, and R&D process templates; when in use, the required industry plug-ins can be deployed one by one through a visual interface according to actual business scenarios; when industry rules or business requirements change, plug-ins can automatically update rule libraries and function modules through an online upgrade mechanism; when processing R&D tasks, plug-ins interact with AI models in real time, automatically inputting industry rules into AI models as analysis constraints to make AI models output results that meet industry-specific characteristics.

5. The method of claim 1, wherein, The monitoring of user operation logs in DevOps tools and the active pushing of content with source identifiers comprises: real-time monitoring of user operation behavior logs in the DevOps tool chain, identifying corresponding R&D scenarios through behavior analysis algorithms; According to the identified research and development scene, matched knowledge assets are retrieved from the structured data asset pool; Through the interface of the DevOps tool chain, the relevant knowledge asset content is actively pushed to the user operation interface, and the source identification is attached; When the user clicks on the pushed content, the source identification can be used to trace back to the original document, quickly obtain complete information, and improve the knowledge acquisition efficiency in the research and development process.

6. The method of claim 1, wherein, The risk closed-loop management includes: Through a risk identification algorithm, potential risks are analyzed, and a risk correlation model is established; The risk correlation model, based on the correlation relationship between historical risk data and whole-process data, can trace back to the associated data of the risk through the unique identifier across the links when detecting a risk in a certain link, and automatically predict the impact range of the risk on subsequent links; The system generates early warning information according to the risk level and impact range, and pushes it to the work interface or message terminal of the relevant role; The system tracks the risk processing state and records the processing process and results, and returns the above risk processing process and result data to the structured data asset pool for optimizing the risk correlation model and further reducing the recurrence rate of similar risks.

7. An apparatus for AI-driven intelligentization of a multi-industry DevOps full process, characterized in that, The device includes: A data asset construction module for interfacing with the DevOps tool chain, extracting whole-process key data, establishing a unique identifier across links and correlating the key data, and constructing a structured data asset pool after processing the extracted data; An AI model architecture module for constructing an AI model layered architecture combining a basic model and an extended model, the basic model layer deploying standardized AI general capabilities, and the extended model layer providing a plug-in interface to automatically achieve different intelligent analysis effects according to different AI task types; An industry plug-in management module for built-in industry characteristic plug-ins, one-key deployment and dynamic update of plug-ins according to business scenarios, so that the plug-ins can interact with the AI model in real time and automatically load corresponding industry rules when processing research and development tasks, and output results conforming to industry characteristics; A knowledge pushing module for monitoring user operation logs in the DevOps tool, identifying research and development scenarios, and retrieving matched knowledge assets from the structured data asset pool, actively pushing content through the tool link interface in the user operation interface and attaching source identification to support one-key tracing back to the original document; A risk closed-loop management module for analyzing potential risks and establishing a risk correlation model based on the unique identifier associated key data, automatically predicting the impact range on subsequent links when detecting a risk in a certain link, generating early warning information and processing suggestions and pushing them to the relevant role, and tracking the processing state to form a closed-loop management and reduce the recurrence rate of similar risks.

8. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a computer program, and the computer program is executed by the processor to implement the method of any one of claims 1-6.

Images