Finance Expert
At Mercor, I worked as a financial AI reviewer responsible for expert-level labeling and annotation of investment analysis outputs generated by large language models. My role focused on rubric-based evaluation, where I systematically classified responses across multiple financial criteria, identified reasoning gaps, inconsistencies, and structural distortions, and provided detailed qualitative annotations to justify pass/fail decisions. This required deep knowledge of portfolio construction, risk metrics, and institutional investment frameworks. In addition to grading model outputs, I contributed to refining evaluation standards by stress-testing prompts and ensuring alignment between expected financial reasoning and model performance. The work went beyond simple tagging; it involved structured analytical annotation, error pattern identification, and cognitive complexity assessment to improve the robustness and reliability of AI systems operating in finance.