Keith User

I’m a highly analytical, detail-oriented AI evaluator and prompt engineer with 5+ years of experience in AI training, data annotation, and content evaluation. I consistently deliver high accuracy and quality by following complex instructions, working independently on remote projects, and meeting stringent deadlines.

Since 2022, I’ve evaluated 5,000+ AI-generated outputs for safety guideline adherence, factual accuracy, and linguistic quality—helping reduce error rates by 15%. I also develop and refine 200+ prompt engineering strategies to elicit targeted responses, improving content relevance and reducing prompt-response cycles by 20%.

I bring rigorous research and quality assurance to every task: I fact-check AI outputs using diverse internet sources to protect data integrity, and I’ve supported large-scale labeling work by annotating 100,000+ data points across text, image, and audio. Whether reviewing content or verifying information, my focus is clear—raise reliability, strengthen model outputs, and maintain dependable standards.