- Red-team LLM-powered systems including chatbots, copilots, RAG pipelines, and AI agents.
- Test for jailbreaks, prompt injection, system-prompt leakage, and tool misuse.
- Write lightweight Python to automate attacks, collect responses, and generate repeatable reports.
- Build and maintain an internal library of prompts, test cases, and regression tests.
- Convert successful attacks into regression tests and clear, reproducible bug reports.
- Track new red-team techniques and integrate them into internal testing frameworks.
- Support GTM teams by producing evidence for customer demos and security reviews.
PythonQA AutomationLLM