About
For more than a decade I’ve built the data systems companies run on: real-time pipelines and billing on billions of events a day, and analytics that bring many sources together into one view. I’ve built AI agents since before agents went mainstream, and run them in production at scale: assistants that have handled millions of conversations over SMS, chat and voice.
In every company I’ve worked at, the people running it wanted the full picture, built from every source they had, not one tool’s view. Now that AI is everywhere, everyone also wants it built on that picture: assistants, workflows and automations.
Building AI agents taught me they’re only as good as the data they work from. That’s the work I do now: one picture you can trust, and the AI tools that run on it, built with your team’s domain expertise.
Experience
Business-critical systems, high-load and running around the clock, across video advertising, insurtech, engineering intelligence and conversational AI.
Rebuilt the AI engine behind its leasing assistant, then unified SMS, chat and voice into one agentic core, including low-latency phone calls and conversation scoring.
3M+ SMS conversations and thousands of voice AI calls
Backed by Y Combinator and a16z · Acquired by VennSet the technical direction for a team of engineers who each owned their work.
An AI agent built on its own data, before agents went mainstream
Backed by Point Nine · Acquired by The Linux FoundationDirected engineering for LFX Insights, which shows the health of open-source projects: who contributes, how active each one is, and how it compares. The team spanned the Americas to Asia-Pacific.
Analytics on 15,000+ repositories and 1.9M contributors
2,000+ member organisationsRebuilt the analytics pipeline to handle billions of events a day in real time, cutting time-to-query from over 20 minutes to seconds, and rebuilt billing to run in real time, counting every event exactly once.
Real-time analytics and billing on billions of events a day
Acquired by ShowHeroes GroupCustom dashboards, A/B testing, RAG systems and AI chatbots, plus classic machine learning: classification, sentiment analysis and face recognition.
Working together
We talk through what you need from your data, and whether I’m the right person to help.
If there’s a fit, you get a scope with dates before any work begins.
You work with me directly, and your team stays in the loop on every decision.