// what i build for clients

Services

Every one of these is something I've shipped to production — not a capability I'd like to try. Pick the one that matches your problem.

Multi-LLM Systems

A model router that picks the right LLM per task — so you stop paying GPT-4 prices for jobs a faster, cheaper model handles just as well.

Learn more

Voice AI Agents

AI that answers and makes real phone calls — handles the full conversation, books the appointment, qualifies the lead, and hands off cleanly when it should.

Learn more

Automation (n8n / Make)

The repetitive 20-hours-a-week ops work — running quietly in the background while your team focuses on the stuff that actually needs a human.

Learn more

RAG & Chatbots

Retrieval that actually returns the right chunk — not a demo that looks great until question #3 breaks it.

Learn more

AI API Integrations

Wiring LLMs, tools, and your existing product together — the unglamorous plumbing that makes an AI feature actually ship.

Learn more

FastAPI LLM Backends

Serving LLMs in production with FastAPI — streaming, auth, rate limits, and observability built in. Not a script that dies under real load.

Learn more

Design to Development

Your Figma file isn’t a website yet. I turn finished designs into clean, pixel-accurate, production-ready code — responsive, maintainable, and ready to ship.

Learn more

SaaS Building

You have an idea and a Notion doc. I turn that into a working, deployed product — scoped tightly, built properly, and ready for real users from day one.

Learn more

AI Consultation

One hour, your idea, and someone who’s actually shipped this stuff in production. Walk away with a clear direction — not more questions.

Learn more

// the north star: your project, shipped

Have an AI project in mind?

Book a free 30-minute consultation. No slide deck — I’ll tell you straight if it’s worth building, and roughly what it takes.

Book a consultation Send a message

Booking 2 projects for Q3 2026