Architect, builder, founder. I'm a Deep Learning Architect at AWS, where I lead production deployments of agentic AI for Fortune 500 companies and startups. On the side I open-source tools for the agent reliability layer.
I lead 30+ GenAI production deployments averaging an 8-week time-to-value. I architect and code agentic AI, RAG, IDP, conversational AI, and browser automation systems. I've developed 15+ reference architectures adopted across the organization, delivered 35+ executive workshops, and helped customers cut operational costs by 40 to 50%.
Ten public projects. The current work is the agent reliability layer: Browser Agent Protocol, DBAR, uSEID, skill-tools, agent-contracts, and agent-pager. The rest is speech AI research from earlier.
I'm also building transparent.fyi, a barcode scanner for food and beverages. It surfaces what the FDA, EFSA, IARC, and peer-reviewed research actually say about every ingredient. 544 ingredient pages so far. Still pre-launch.
I started it for myself. I care a lot about what I eat, and the nutrition label hides the things that matter. I wanted transparency, and a faster way to read an ingredient list.
Senior ML engineer at Numerade Labs. Senior ML scientist at Spectrum Labs, where I scaled a system to 12+ languages processing billions of daily messages. Research engineer at Comcast Applied AI and Indiana University Bloomington. I have three published papers on speech AI; two won best-paper awards at INTERSPEECH 2021 and IEEE ICASSP 2021.
Pickleball most weekends. The Katy Trail when it's cool out. Uchiba on a quiet Tuesday.