At FHLB San Francisco, I lead the credit and collateral data platform and delivered two governed AI agents to production: a credit memo agent used by all eight analysts and a policy navigator with 70% weekly use. I also embedded credit scoring outputs, sensitivity analysis, and early-warning signals in underwriting decisions, cutting quarterly review time by 50%.
I partnered with Risk, Compliance, and Technology on AI guardrails and built a single source of truth integrating S&P, Bloomberg, and other providers. After validating the approach with an MVP, I reduced the data error rate from 5.2% to 0.1%, cut manual effort by 80%, and increased collateral throughput by 50%.
At Moody’s Analytics, I led a team of eight across a $50M+ ARR lending and risk portfolio. Earlier, I converted a spreadsheet prototype into an API-first SaaS product that reached 60+ customers and $10M+ ARR, and I bring a quantitative background in credit model validation, a CFA charter, and hands-on experience with SQL, Python, Power BI, Microsoft Copilot, and Claude Code.

