I'm an engineering leader who builds AI-native platforms and helps teams ship the way one AI-leveraged builder does — turning AI from a few power users into a whole-team operating model.
Most of my work lives in the messy middle where ideas become shipped systems: platform architecture, CI/CD, cloud infrastructure, and the guardrails, context, and agentic workflows that let teams move in days instead of quarters — without the wheels coming off.
- AI-native development practice — guardrails, review/security gates, reusable context and agent patterns, team training and adoption
- Platform engineering — monorepo architecture, shared APIs, CI/CD, AWS serverless infrastructure
- Engineering leadership — taking vision and turning it into shipped product; mentoring builders; modernizing how teams work
- Turning legacy and ambiguity into momentum — within real budget and organizational constraints
🪺 Roost — a modular monorepo framework for personal and business use. Plugins for AWS CI/CD and infrastructure build-out with cost/resource tagging, email service integrations, and the platform patterns I've found worth reusing. Open source, built from the ground up. (link coming soon)
Going deep on language model fundamentals via Stanford's CS336 (Language Modeling from Scratch) — because understanding the model layer from first principles makes you better at steering it.
TypeScript · Next.js · React · Node.js · GraphQL · AWS (Lambda, RDS, Cognito, SST) · PostgreSQL · Turborepo · Claude · Cursor
I care more about solving the problem than about any particular tool. The interesting question isn't "what's your stack" — it's "can you turn chaos into something that works?"