AI Beta Tester & LLM Evaluator
Performed red-team testing on LLM systems to identify policy bypass vulnerabilities. Designed adversarial prompts to test content moderation and safeguard effectiveness. Evaluated model responses for safety, hallucinations, bias, and logical consistency. Created and refined system prompts for persona-based AI simulations. Conducted text and image annotation using Label Studio. Followed strict annotation guidelines and quality standards. Contributed to AI model training and alignment improvements