AI Engineeringfor Unstructured Financial Data

I'm a hybrid AI engineer and quantitative modeler with a background that bridges fintech infrastructure, machine learning research, and full-stack AI systems.

At Katapult, I helped build and scale real-time underwriting pipelines that powered over $250M in annual originations. My work combined applied machine learning, fraud detection, and dynamic pricing — with direct financial impact. As Risk Infrastructure Lead, I also led the buildout of production tooling and model deployment architecture, driving both velocity and rigor across the ML lifecycle.

Alongside industry work, I've contributed to applied ML research during my M.S. at NYU Courant, with projects spanning natural language processing, vector representations, and optimization. That academic foundation — in linear algebra, probability, and numerical methods — still informs how I design and interpret models in high-stakes, signal-rich environments.

Today, I'm focused on systematic signal extraction from unstructured data — building full-stack applications that turn transcripts, reviews, and filings into real-time trading intelligence. I'm fluent in modern vector DBs, LLM pipelines, retrieval-augmented generation (RAG), and custom API integrations.