Case 3 of 4 · 2025 – now
Exploratory testing, automated with AI
An LLM-powered framework that replaced hours of manual exploratory testing on every release.
My role QA Lead; built the framework
Context
Every release needed hours of manual exploratory testing across backend services.
Strategy
Use a large language model to drive functional and exploratory checks, and generate service documentation in the same CI/CD pipeline.
What I did
- Built an AI-powered framework for functional and exploratory testing.
- Integrated LLM-based documentation generation into CI/CD across 6 services.
Results
| Metric | Before | After |
|---|---|---|
| Manual exploratory testing per release | Hours | Automated |
| Services with generated docs | — | 6 |
Stack
- LLM
- Python
- GitLab CI