AI+python-based qa automation auditing method and system

Abstract

The application discloses a kind of based on AI+Python's QA automatic auditing method and system, the method includes each inspection item in QA auditing inspection sheet, is audited by auditing tool, and software project code and software version number to be audited are input in auditing interface;Automatic auditing tool and other platforms interact, pull relevant information in each database;The information pulled from each platform is judged according to rules, and the auditing conclusion is obtained;The auditing conclusion is normalized output, forms QA automatic auditing report, and each inspection clause is given auditing conclusion one by one, and the specific problems existing are shown;The system includes auditing tool module, interactive platform module, judging module and output module.The application not only saves the time that QA personnel manually carries out process auditing, but also improves the coverage and correctness of process auditing.

Description

Technical Field

[0001] This invention relates to the field of quality management of aero-engine control software, and in particular to an AI+Python-based automated QA audit method and system. Background Technology

[0002] During software development, software QA needs to conduct quality assurance audits throughout the entire R&D lifecycle, involving multiple systems such as review, configuration management, and version control. Currently, the lack of unified interfaces and automated data collection mechanisms between these systems forces auditors to frequently switch between different platforms, manually collect and compare data, resulting in low efficiency. Furthermore, due to time and resource constraints, when the scale of products to be audited is large, sampling is often the only option, which is prone to omissions; and manual audits themselves are inherently difficult to completely eliminate human error. Therefore, it is necessary to introduce automated audit methods to improve audit efficiency, achieve full coverage of audit content, reduce omissions, and minimize human error. Summary of the Invention

[0003] Purpose of the invention: The purpose of this invention is to provide an AI+Python-based automated QA audit method and system to save QA personnel time in manually conducting process audits and improve the coverage and accuracy of process audits.

[0004] Technical solution: The AI+Python-based automated QA review method of this invention includes the following steps:

[0005] S1. Solidify each check item in the QA audit checklist in Python syntax, and combine the Python source files to form a coded audit tool;

[0006] S2. Conduct the review using the review tool. Enter the software project code and software version number to be reviewed in the review interface.

[0007] S3, the automatic review tool interacts with other platforms, matching software project code and software version number to retrieve relevant information from various databases;

[0008] S4. Analyze the information retrieved from various platforms according to the rules and draw a review conclusion;

[0009] S5. Standardize the output of audit conclusions to generate an automated QA audit report, providing audit conclusions for each inspection item and showing the specific problems found.

[0010] Furthermore, step S1 employs a Python-based automated QA auditing method using terms as code.

[0011] Furthermore, in step S1, each check item in the QA audit checklist is solidified in Python syntax. Each check item forms a Python source file, and the Python source files are combined to form a coded audit tool. Each check item corresponds to one Python file.

[0012] Furthermore, the review process in step S2 includes inputting the software project code name and version number to be reviewed. The inspection tool matches the input information and interacts with systems such as configuration management, review, and project management based on the matching information to obtain relevant documents and code information. Then, Python+AI is used to conduct process and product reviews.

[0013] Further, step S3 includes:

[0014] S31. Retrieve the relevant code tags, baseline information, and creation time from the project management platform database;

[0015] S32. Retrieve the relevant review object, review version, review time, and review conclusion from the review system database;

[0016] S33. Obtain the relevant file, file version information, and file creation time from the configuration management system.

[0017] Furthermore, step S4 includes calling the AI ​​big model, inputting the "review requirement prompts" into the big model in combination with the actual business situation to complete the product review and return the AI ​​review results, and QA confirming the questions raised by the AI ​​review.

[0018] Furthermore, step S5 includes checking the software's main responsibility scan, checking the project baseline, remaining defects, and checking the integrity of the review role.

[0019] The AI+Python-based automated QA review system of this invention includes:

[0020] The audit tool module is used to solidify each check item in the QA audit checklist in Python syntax and combine the various Python source files;

[0021] The interactive platform module is used to interact with the automated review tool, matching software project code and software version number to retrieve relevant information from various databases;

[0022] The judgment module is used to perform rule-based judgments on the information retrieved from various platforms and draw audit conclusions.

[0023] The output module is used to standardize the output of audit conclusions, generate an automatic QA audit report, provide audit conclusions for each inspection item, and display the specific problems found.

[0024] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages:

[0025] Experience is solidified without loss: Each process check item in the existing QA checklist is independently mapped to an independent Python code unit, realizing "item as code" and fully inheriting the original experience;

[0026] AI Intelligent Review: By calling upon a large AI model, the business knowledge base is deeply integrated with existing QA experience to automatically conduct product reviews and return AI review conclusions, and generate review reports for QA manual confirmation or rapid review.

[0027] One-click, extremely simple operation: Users only need to enter "project code + version number", and the tool will automatically complete data retrieval, rule determination and report generation, with zero human intervention throughout the entire process;

[0028] Easy to maintain and expand: Each process review clause corresponds to a single *.Py file. Adding, deleting, or modifying clauses only requires modifying this file, resulting in low maintenance costs and flexible expansion. Attached Figure Description

[0029] Figure 1 This is a flowchart illustrating the interaction between the automatic review tool of this invention and other systems.

[0030] Figure 2 This is a configuration interface diagram for the automatic review tool of this invention;

[0031] Figure 3 This is a report generated by the automatic review tool of this invention. Detailed Implementation

[0032] The technical solution of the present invention will be further described below with reference to the accompanying drawings.

[0033] The existing QA audit checklist items can be automated to a certain extent using Python and AI. Users only need to provide the software project code name and version number to perform automatic audits and generate audit reports.

[0034] 1. For process audits, a Python-based automated QA audit method called "Clause-As-Code" is adopted. Each check item in the QA audit checklist is solidified in Python syntax, with each check item forming a Python source file (*.Py). The Python source files are combined to form a coded audit tool. Each check item corresponds to a Python file, which is easy to maintain.

[0035] 2. For the product review that was previously conducted using a sampling method, we now call upon the AI ​​big model, input the "review requirement prompts" into the big model based on the actual business situation, complete the product review, and return the AI ​​review results. QA only needs to confirm the questions raised by the AI ​​review.

[0036] When using the audit tool, users enter the software project code and version number to be audited. The audit tool matches the input information and interacts with configuration management, review, project management and other systems based on the matching information to obtain relevant documents and code information. Then, it uses Python+AI to conduct process and product audits.

[0037] The interaction process between automated review tools and other systems, such as Figure 1 As shown.

[0038] The AI+Python-based automated QA review method described in this invention includes the following steps:

[0039] Step 1: Solidify each check item in the QA audit checklist into Python syntax. Each check item forms a Python source file (*.Py). Combine the Python source files to form the QA automated audit tool.

[0040] Step 2: In the audit tool configuration interface, enter the software project code and software version number to be audited. The configuration interface is as follows: Figure 2 As shown.

[0041] Step 3: The automated review tool interacts with other platforms, matching software project code and software version number to retrieve relevant information from various databases:

[0042] 3.1) Retrieve relevant code tags, baseline information, creation time, etc. from the project management platform database;

[0043] 3.2) Retrieve the relevant review objects, review versions, review times, review conclusions, etc. from the review system database;

[0044] 3.3) Obtain the relevant files, file version information, file creation time, etc. from the configuration management system.

[0045] Step 4: Analyze the information retrieved from various platforms according to the rules and draw a review conclusion;

[0046] Step 5: Standardize the output of the audit conclusions to generate an automated QA audit report. Provide audit conclusions for each inspection item and display specific issues. The audit report format is as follows: Figure 3 As shown.

[0047] The AI+Python-based automated QA review system of this invention includes:

[0048] The audit tool module is used to solidify each check item in the QA audit checklist in Python syntax and combine the various Python source files;

[0049] The interactive platform module is used to interact with the automated review tool, matching software project code and software version number to retrieve relevant information from various databases;

[0050] The judgment module is used to perform rule-based judgments on the information retrieved from various platforms and draw audit conclusions.

[0051] The output module is used to standardize the output of audit conclusions, generate an automatic QA audit report, provide audit conclusions for each inspection item, and display the specific problems found.

Claims

1. A QA automated review method based on AI+Python, characterized in that, Includes the following steps: S1. Solidify each check item in the QA audit checklist in Python syntax, and combine the Python source files to form a coded audit tool; S2. Conduct the review using the review tool. Enter the software project code and software version number to be reviewed in the review interface. S3, the automatic review tool interacts with other platforms, matching software project code and software version number to retrieve relevant information from various databases; S4. Analyze the information retrieved from various platforms according to the rules and draw a review conclusion; S5. Standardize the output of audit conclusions to generate an automated QA audit report, providing audit conclusions for each inspection item and showing the specific problems found.

2. The AI+Python-based automated QA auditing method according to claim 1, characterized in that, Step S1 employs a Python-based automated QA auditing method using terms as code.

3. The AI+Python-based automated QA review method according to claim 1, characterized in that, Step S1 solidifies each check item in the QA audit checklist in Python syntax. Each check item forms a Python source file, and the Python source files are combined to form a coded audit tool. Each check item corresponds to one Python file.

4. The AI+Python-based automated QA auditing method according to claim 1, characterized in that, The S2 review process includes inputting the software project code and version number to be reviewed. The inspection tool matches the input information and interacts with configuration management, review, project management and other systems based on the matching information to obtain relevant documents and code information. Then, Python+AI is used to conduct process and product review.

5. The AI+Python-based automated QA auditing method according to claim 1, characterized in that, Step S3 includes: S31. Retrieve the relevant code tags, baseline information, and creation time from the project management platform database; S32. Retrieve the relevant review object, review version, review time, and review conclusion from the review system database; S33. Obtain the relevant file, file version information, and file creation time from the configuration management system.

6. The AI+Python-based automated QA auditing method according to claim 1, characterized in that, Step S4 includes calling the AI ​​big model, inputting the "review requirement prompts" into the big model in combination with the actual business situation to complete the product review and return the AI ​​review results, and QA confirming the questions raised by the AI ​​review.

7. The AI+Python-based automated QA auditing method according to claim 1, characterized in that, Step S5 includes checking the software's main responsibility scan, checking the project baseline, identifying remaining defects, and checking the integrity of the review role.

8. A QA automated review system based on AI+Python, characterized in that, include: The audit tool module is used to solidify each check item in the QA audit checklist in Python syntax and combine the various Python source files; The interactive platform module is used to interact with the automated review tool, matching software project code and software version number to retrieve relevant information from various databases; The judgment module is used to perform rule-based judgments on the information retrieved from various platforms and draw audit conclusions. The output module is used to standardize the output of audit conclusions, generate an automatic QA audit report, provide audit conclusions for each inspection item, and display the specific problems found.

Images