Engineering leader with 15+ years building products across AI, security, and infrastructure. Currently Member of the Technical Staff at OpenAI, where I work on Promptfoo and agent security. Previously co-founded Promptfoo, built it into an AI security platform, and sold it to OpenAI. Stanford MS, YC alum, previously VP Engineering.
Promptfoo started as a developer-first eval tool and grew into a platform
for AI security, red-teaming, and compliance. We scaled it to more than 350,000 developers, 130,000
monthly active users, and teams at more than 25% of the Fortune 500 before selling the company to OpenAI.
Owned company-wide technical strategy and product roadmap across open-source adoption, commercial product, and research investments.
Co-led the company through its sale to OpenAI in March 2026.
Helped turn Promptfoo from an open-source eval tool into an enterprise AI security product used by large companies and frontier labs.
Built and scaled the company from founding to 23 people across engineering, GTM, and operations.
Led technical due diligence and fundraising through $23.4M in venture financing from Insight Partners and Andreessen Horowitz.
Led development of core evaluation framework, vulnerability scanning, static analysis, and automated red-teaming capabilities.
Smile Identity provides ML-powered identity verification APIs used by banks, fintechs, and
telcos across Africa. Promoted from Director to VP Engineering to VP Engineering & Head of AI within one year,
leading teams building APIs that now process hundreds of millions of identity checks.
Owned engineering org design, headcount planning, and hiring across backend and ML teams, growing to 20+ engineers.
Transformed engineering velocity from weekly releases to continuous deployment, migrating to TypeScript, adding tests in CI, and leading ceremonies.
Re-architected inference APIs on AWS Lambda, scaling from 1,000 to 1M+ users per day and cutting job time from 30+ seconds to 7 seconds.
Built computer vision pipelines for liveness detection certification.
Pitched, designed, and shipped a fraud detection product using 1-N facial recognition with embeddings and vector search.
Arthena was a quantitative art investment platform backed by Anthemis,
Foundation Capital, and Y Combinator.
Built Arthena from idea to acquisition by Masterworks in 2023.
Co-founded Arthena and led technical strategy; sat on board and led fundraising, including debt financing for a 9-figure investment vehicle in the auction guarantee market.
Built and managed a cross-functional team of 20 engineers, data scientists, and analysts.
Built data pipelines, quantitative research tools, and visualization systems to scale investment research and augment analyst decision-making.
Developed valuation models on irregularly-sampled time series using graph embeddings, probabilistic forecasting with calibrated prediction intervals, and online learning with walk-forward validation.
Designed micro-service architecture for data collection, feature engineering, backtesting, and reporting.
Matroid is a computer vision platform for creating and deploying detectors, now funded by
NEA and Accel. I co-founded
the company and built the initial product through our first fundraise.
Defined company vision and product strategy.
Architected and built the initial detector platform for identifying objects, events, and patterns in video.