Thales P.

Currently leading high-impact initiatives focused on AI evaluation, bridging the gap between complex data requirements and model performance. My role has evolved from operational data labeling to designing Gold-Standard Benchmarks and managing large-scale Quality Assurance (QA) workflows dedicated to the evaluation of frontier Large Language Models (LLMs). I bring a rigorous, logic-driven perspective to AI training. My focus is on evaluating reasoning capabilities, identifying edge cases, and optimizing resource allocation within data pipelines, ensuring that models are trained on high-quality data.