Inspiration Manual API testing and debugging is repetitive and slow. As a QA automation engineer, I often spent hours checking logs, payloads, and HTTP errors to find the root cause. I wanted a faster way to turn raw failure data into clear insights using AI — especially with the new Gemini 3 API and Google’s official GenAI SDK. What it does This tool automatically analyzes failed API tests using Gemini 3 and: Finds the root cause of failures Explains errors in simple language Suggests possible fixes It helps teams debug faster and deliver more reliable APIs. How we built it We run API tests, collect failure data (logs, payloads, status codes), and send it to Gemini 3 using the official google-genai SDK. Gemini then returns clear, actionable insights. Challenges The hardest part was formatting the failure data so the AI could understand it properly. Also, we migrated from the old google-generativeai SDK to the official google-genai SDK. What’s next Web dashboard for failure reports Browser testing support (Selenium/Playwright) CI/CD integration for continuous monitoring
AI-powered API Testing & Failure Analysis
From failing tests to clear fixes, AI-driven QA automation.
Updates
This project now uses Gemini 3 to automatically analyze failed API tests, explain errors in plain language, and suggest fixes. It runs API tests with Pytest, captures failure logs, and returns intelligent insights via a Flask API. Next steps include adding a UI dashboard, CI/CD integration, and browser testing support.
Leave feedback in the comments!
Log in or sign up for Devpost to join the conversation.
Log in or sign up for Devpost to join the conversation.