Handshake AI — AI Generalist reviewer & AI Generalist Trainer (AI training / RLHF-style feedback)
Served as an AI Generalist reviewer and AI Generalist Trainer, providing human feedback to improve foundational model quality and compliance. Assessed LLM outputs across factuality, logical reasoning, code quality, and adherence to safety guidelines to support training and evaluation. Authored and executed prompt engineering strategies to stress-test constraints, uncover edge cases, and reduce hallucinations. • Provided RLHF-style human feedback and detailed annotations for fine-tuning • Improved response quality and policy compliance by tracking impact metrics • Analyzed error patterns with teams to refine training datasets • Integrated evaluation insights into model assessment workflows.