Software Engineering & Data Science Consultant – RLHF/AI Training
As a Software Engineering & Data Science Consultant at Outlier AI, I contributed directly to RLHF pipelines by evaluating and ranking outputs from large language models. My work involved authoring and reviewing high-complexity prompts and reference solutions in Python, SQL, and ML, establishing high-quality benchmarks for LLM improvement. I identified and documented systematic failure modes, providing actionable, structured feedback to enhance model performance. • Evaluated and ranked computer code outputs for LLM refinement. • Authored and reviewed complex prompts and solutions for use in AI training datasets. • Provided feedback to target logical errors and efficiency issues in generated code. • Directly contributed benchmark data and evaluations for the development of next-generation LLMs.