AI Engineer – AI Rubric Creation and Evaluation
In this role, I created rubrics and trained AI systems to generate Python and SQL code based on user input. My responsibilities included evaluating and verifying AI trainer inputs to ensure data quality, and refining model loss and gradient functions during training loops. I was tasked with improving AI output quality by designing and adjusting rubrics for technical accuracy. • Developed benchmarks and criteria to assess AI-generated code outputs. • Evaluated user-generated code to enhance dataset complexity and correctness. • Provided corrective feedback for AI and trainer submissions, strengthening Python and SQL code outputs. • Utilized internal/proprietary tools provided by Telus to guide the training and evaluation process.