Subject Matter Expert (AI & Computer Science) — AI-generated code auditing, benchmarking, and rubric-based evaluation
Performed rubric-based evaluation and scoring of complex multi-turn, AI-generated source code for correctness and efficiency. Reviewed execution efficiency, structural accuracy, and adherence to coding standards while generating benchmarked outcomes for model improvement. Ranked model code outputs using criteria focused on time/space complexity, edge-case vulnerability handling, and memory allocation. • Audited multi-turn code and architectures for execution efficiency and structural accuracy • Developed logical reasoning programming problem sets for benchmarking and fine-tuning • Scored and ranked outputs using precision rubrics for complexity and robustness • Debugged and optimized back-end data pipelines and automated grading scripts for RLHF workflows