AI Tool
for Testing
Web APIs
Test faster, detect issues and defects in code earlier, and release web applications more frequently with our AI tool for testing APIs.
How Does Akvelon’s AI-driven API Testing Tool Work?
- Automate the creation of test cases for APIs based on analyzed documentation and API responses
- Auto-create test scenarios for a range of valid and invalid inputs to simulate how real users may interact with your project’s API
- Generate tests for evaluating API response codes and content in received JSON data and cover written test cases
- Identify edge cases — scenarios that are at the boundaries of input parameters, unusual user behavior, errors, or exceptions
Ensure Your Application’s APIs Function as Expected in Four Steps
The application’s API integrates into the tool. A JSON file with API is received as the input. The tool can use a single API or a collection of APIs.
The application’s API integrates into the tool. A JSON file with API is received as the input. The tool can use a single API or a collection of APIs.
Test engineers review the file and change it if needed by adding new test cases, deleting, or updating existing ones.
The API testing tool creates the prompt to write test scripts based on previously generated test cases. Executable test scripts form a Python file.
Technology stack
OpenAI API, Python, Python ast module, Pytest, Swagger (v2 and v3), Swagger Petstore
Impact on Testing Speed and Cost
- The solution creates 20 executable tests for an API endpoint in 1:38 min, while a middle SDET engineer needs up to 20 hours to complete this task.
- The tool needs 13.6 hours to check 50 endpoints with 20 tests, which would usually take a human specialist 600 hours to complete.
- Using an LLM to generate test cases and scripts would cost $0.064 per endpoint or $3.20 for 50 endpoints. The average SDET’s work costs significantly more.
Based on experiments and feedback from clients to whom we introduced the AI tool for testing, we estimated that it can reduce a test engineer’s workload by approximately six times. So, the count for writing test scripts for one API is now in minutes instead of hours. Thanks to increased testing efficiency, tech companies can deliver high-quality software to their customers faster.
Nail Shakirov
Director Of Engineering at Akvelon
Nail Shakirov
Director Of Engineering at Akvelon
Ready to reap the benefits of using AI for software testing? Reach out to us for more details!
Have a question?
The tool generates functional tests to ensure that an API gives expected outputs for specific inputs and handles valid and invalid user inputs and edge cases. These tests can help simulate real-world usage scenarios of the API.
With our AI tool for testing, you can rest assured that your data will be safe. Our tool only needs Swagger API descriptions to create tests — that’s why there is zero risk of Personal Identifiable Information (PII) exposure. Your customers’ data remains untouched since it works solely with JSON metadata.
Currently, we don’t charge our clients for using the tool. It’s a value-added benefit we provide as part of our tech partnerships. In addition, developing and maintaining this solution lets us advance our expertise in leveraging AI for software testing.