Senior Portfolio Manager (BlackRock) — AI finance model evaluation, rubric/benchmark design, and structured feedback for model improvement
TextEvaluated AI-generated financial and market commentary for factual accuracy, logical consistency, analytical depth, risk awareness, and practical usefulness in investment contexts. Designed evaluation frameworks to assess model outputs on finance tasks such as valuation assumptions, market analysis reasoning, portfolio strategy, and overall forecasting discipline. Identified recurring model failure modes including unsupported assumptions, weak risk framing, shallow market interpretation, calculation errors, and overconfident recommendations. • Produced structured written feedback to improve model behavior and evaluation consistency. • Created rubrics and benchmarks to standardize performance assessment across finance-related tasks. • Reviewed trading logic, position sizing concepts, market impact reasoning, and execution-related decision-making outputs. • Calibrated risk/reward assessments by challenging model assumptions relevant to portfolio and trade outcomes.