AI Model Evaluator at Outlier AI

Wrote code examples, debugging walkthroughs, and explanations in JavaScript, Python, and CSS to train and evaluate LLMs on problem-solving accuracy and documentation quality. Identified and corrected logical errors in AI-generated code, delivering structured feedback that measurably improved model reasoning reliability and output consistency. Reviewed and assessed AI-generated outputs across JavaScript, Python, and CSS against industry standards, flagging inconsistencies and edge cases to improve model behaviour. Maintained high evaluation quality and throughput in a fast-paced async remote environment, consistently meeting turnaround targets across a high weekly volume of submissions.