At Handshake AI solution, I evaluated and refined LLM outputs, writing prompts to test model behavior across interactive, visual, and adaptive scenarios.
At Toloka AI, I analyzed search results and AI-generated responses for relevance, clarity, and factual correctness. I sustained a 98% quality rating for consecutive months and identified edge cases that contributed to refining model behavior and guidelines.
I also classified and labeled complex LLM datasets at AI Data Solutions Inc., placing in the top 5% of annotators. Outside that work, I built React projects including CVAI Tailor, a resume tool with LLM integration and a live job-match score.
