About Me

I'm an AI engineer and consultant who builds production AI systems for real businesses.

My path here was anything but linear. I studied mechanical and electrical engineering as an undergrad, got a master's in computer science, spent over a decade in software engineering, then did a PhD in Bayesian Econometrics, mostly part-time while shipping software. That degree pulled me into data science, machine learning, and a decade of computational and statistical research before I jumped into the current wave of AI. Oddly, the thing that made me pull the trigger wasn't GPT-3. It was Stable Diffusion.

I was building deep learning models at Pristine when ChatGPT came out and GPT-4 hadn't dropped yet. Then I joined Sixfold.AI as their first AI engineer, where I spent two and a half years building an insurance copilot and learning hard lessons about what it actually takes to ship AI in production. I left Sixfold in May 2025 to start my own consulting practice.

I work from a first-principles perspective and a simple conviction: build the system around the LLM, not on top of it. Models are powerful but unreliable components. The real engineering is in the structure you put around them: the boundaries, interfaces, evals, and feedback loops that make the whole thing reliable.

My methodology rests on four pillars: infrastructure that makes experimentation cheap, deep domain knowledge that tells you where to draw the boundaries, evals that turn guessing into engineering, and a data flywheel that makes the system better over time.