Over the past 27 years I've built products with companies of every size, scrappy startups to established enterprises, shipping things that actually scale. For the last few years that work has centered on AI: systems that hold up in production and earn their place in a real business. Anyone can ship a demo.
The deepest of those years went into telecom: more than fifteen of them on phone systems, SMS, and the messy plumbing that connects them. I worked with Asterisk and FreeSWITCH, and I got into Twilio early enough that I literally wrote the book on it. I moved into WebRTC once video landed in the browser, and I ship both sides of that world: video platforms people could join from a browser tab, and dialer systems that have to place calls reliably at volume.
Real-time systems are unforgiving in a way most software isn't. When a call drops or a stream stutters, the person on the other end knows right away. Years of that taught me to think in latency, retries, and failure states, which turns out to be most of what matters when you put AI into production too.
As an AI Product Engineer, I work backwards from what the product should be, then build it end to end: frontend, backend, infrastructure, and the AI layer woven through all of it. I'm at my best when handed a rough idea and asked to make it real. That means knowing where AI creates genuine leverage and, just as importantly, where it doesn't.
In practice, that shows up as AI automation, integrating AI into the parts of a business where it compounds: lead generation, internal tooling, content pipelines, and the workflows that quietly eat a team's time. My stack here includes Claude, ChatGPT, Cursor, Hermes, MCP servers, n8n, and LangChain, among others.
And as a Fractional CTO, work I take on through Data McFly, I step into the technical leadership gap, working with founders and leadership teams to set engineering direction, make the architectural calls, and build the systems and teams that let a company scale without a full-time hire. Defining your stack, leading a migration, hiring your first engineers, or translating business goals into technical reality: the same hands-on approach, one level up. It runs on the 70/30 Method: automate the repeatable 70%, keep the decisive 30% with people.
On the technical side: JavaScript across the board (Astro, React, Datastar, Remix, Next.js, Vue, Nuxt), plus PHP, Rust, Go, and Python, with deep experience in headless architecture (Directus, Strapi, Sanity, WordPress, Shopify) and infrastructure from greenfield to legacy migration.
If any of this sounds like what you need, get in touch. I'm always happy to talk through a problem, whether it turns into work together or not.