Jackson User

I’m an AI Training Specialist and LLM Code Evaluator focused on producing highly accurate, well-documented labeled datasets across image, video, and text modalities. With 3 years of hands-on experience, I deliver training data that supports complex technical, algorithmic, and medical domains.

I evaluate, audit, and rank thousands of AI-generated responses for structural correctness, logical accuracy, and instruction-following—so the outputs are RLHF-aligned and ready for fine-tuning pipelines. I’ve completed 3,000+ advanced annotation tasks (including polygon segmentation, keypoint mapping, and multiclass classification) and consistently maintained an accuracy rating above 95% across multi-layer audit reviews.

I also bring full-stack strength to the work: I’ve shipped production-ready full-stack features using Python Flask and TypeScript and built an AI-powered chat assistant with prompt engineering tests and RLHF quality refinements. Through asynchronous collaboration with Slack and Notion, I document anomalies, align rubrics, and run QA with clear bug documentation and acceptance criteria.