Turning the privacy work in the book into a working prototype. No company formed — yet.
assets/portrait.jpg
A short bio.
I'm an applied AI/ML engineer turned architect. I've spent the last several years shipping intelligent software — from classical ML pipelines to retrieval, agents, and fine-tuned open-weights models. Along the way I wrote Machine Learning to Gen AI Agents, a field guide for people who'd rather build than slide-deck.
Today I prototype the tools the book describes — chief among them Avarna Privacy Gateway, a POC for keeping AI workloads private by default. None of these are companies yet; all of them are working proofs.
A short journey, in landmarks.
Four cards. The agent adds new ones as new things ship.
A field guide from foundations to agents. Audio in production.
From classical ML pipelines to retrieval, agents, and fine-tuned open-weights.
Backend, data, distributed systems — the unglamorous stack AI rests on.
Let's build something trustworthy.
Avarna POC walkthroughs, book inquiries, speaking, teaching, or a good AI conversation — my inbox is open.