Freelance - AI Data Trainer at DataAnnotation Remote
Evaluated and compared AI model responses to assess accuracy, reasoning quality, and improvement areas. Created complex, challenging prompts to probe model limitations and identify failure scenarios. Provided detailed feedback based on observed performance and gaps in the responses. • Response accuracy evaluation • Reasoning and coherence quality assessment • Prompt-based testing for edge cases • Identification of scenarios where the AI fails