I build AI systems that make businesses money.

AI systems that do one of three things: bring in more customers, raise what each one is worth, or cut the hours your team spends on manual work. In practice that means lead qualification and follow-up, onboarding and CRM automation, internal knowledge assistants, automated reporting, and safe connections between AI and the tools you already run.

It starts with an audit. I map how work actually moves through your business, then rank what is worth automating by hours saved against build cost. We scope one system with a measurable outcome, I build and deploy it, and you get the source code and the documentation. Most clients move to a monthly retainer after that, so the system stays monitored, tuned and extended.

Most projects fit one of my four fixed-price packages, and each one shows its price and delivery date up front on the services page. Anything bigger gets a fixed quote before work starts, so there are no hourly surprises.

TypeScript and Python for the systems themselves. Claude and OpenAI for the AI layer. Trigger.dev for durable workflows that survive failures and retries. Postgres with pgvector for grounded retrieval. n8n or Make where a visual tool genuinely fits. I integrate with the tools you already pay for rather than replacing them.

Credentials stay server-side and never reach the model. Access is scoped, so a system can only read the data it is meant to. Everything can run on your own infrastructure, and every decision the system makes is logged and auditable.

Book a call and bring one workflow that is eating your team’s time. In 30 minutes you will know whether it is worth automating, what it would take, and roughly what it would cost. You get a written summary either way.