Services

I work with engineering teams that need to ship AI systems in production, not just in demos. The work tends to fall into three shapes.

RAG and AI agent implementation

Most AI projects stall at the prototype stage. The notebook works, the demo wins the meeting, the system never reaches real users at scale. The gap between “this works on my laptop” and “this runs reliably for ten thousand users a day” is wide and full of unfun engineering. That's the gap I work in.

Concretely, I help teams build:

• Production-grade RAG pipelines with proper evaluation
• AI agent orchestration (multi-agent, A2A, tool use)
• Embedding services that survive a deploy
• The boring infrastructure around the AI: monitoring, fallbacks, cost control, observability

Recent work includes the orchestrator agent for the Google + BBC AI Agents demo at IBC2025, which won the Broadcast Tech Innovation Award.

Get in touch if you're building something specific.

AI engineering advisory

Sometimes you don't need someone to build it; you need a senior engineer to validate the architecture, evaluate vendors, or decide build-vs-buy. I do shorter advisory engagements for that, typically a handful of hours per week over a defined window.

Common scenarios:

• Deciding whether to use a managed vector database or self-host
• Evaluating whether RAG, fine-tuning, or long-context is the right approach
• Sanity-checking an AI roadmap before a board meeting
• Pressure-testing claims in a vendor pitch

Book a call to scope it.

Speaking and writing

I speak and write about practical AI engineering. Past topics have included AI agent orchestration, retrieval-augmented generation, and the gap between AI demos and production systems. I'm currently writing Retrieval-Augmented Generation: An Engineer's Guide to Building RAG Systems with Your Own Data.

Invite me to speak if your meetup, conference, or internal session has a topic that fits.