AI Training & Evaluation – Project Diamond (Handshake AI)
For Project Diamond (Handshake AI), I annotated and evaluated AI-generated responses for accuracy and guideline adherence. My work involved assessing AI responses to ensure consistency and high-quality outputs within conversational applications. I regularly used evaluation protocols to enhance model reliability and user experience. • Evaluated and rated thousands of AI-generated text responses. • Ensured responses met detailed guidelines on correctness and tone. • Identified systematic issues and provided suggestions for improvement. • Maintained comprehensive records and reported analytics to project leads.