01 — Start
"Where does AI actually fit, and is what I have any good?"
Before anyone writes a line of production code, you need a straight answer. Where AI earns its place in your organisation, or whether the system you already built will hold up once real traffic hits it. A short, bounded piece of work that leaves you with a decision you can defend.
3–5 days
What AI can realistically do for your organisation and where it earns its keep. A structured assessment with honest recommendations and a written deliverable you can act on this quarter. If the output can't be used, you get your money back.
1 week
A review of an AI system you already have: prompts, retrieval pipeline, architecture. What works, what will break under load, and what to fix first.
1 week
A deep read of your prompt architecture: drift, coverage, edge cases, and the gap between a demo that impresses and a system that survives production. The story-engineering layer is assessed alongside the technical one.
02 — Build
"I know what I need. I want it built and handed over."
The work that takes a system from idea to something running in production. I build it with your engineers in the room, so the code and the understanding to run it both stay once I leave. No black boxes, no dependency on me afterwards.
3–5 weeks
Your first real AI integration into a business process. Knowledge retrieval, document processing, structured-output assistants. Built alongside your team so they own it after delivery, with the engineers in the room the whole way.
4–6 weeks
Memory-durable agent infrastructure built on the same substrate I run my own practice on. Persistent context, multi-step reasoning, tool integration. Delivered as working code your team understands.
4–6 weeks
A multimodal production pipeline: text to audio to video, with character identity held across frames, voice consistency, caption timing, and A/V synchronisation. Transfers to marketing content, training material, and internal communications.
1 day
One day. I build a working prototype of your real problem in front of your team while I narrate every decision. No follow-on commitment required. The team leaves with code and understanding.
03 — Sustain
"I want a senior AI perspective on call."
Once something is running, the questions keep coming: a vendor to weigh, an architecture call to make, a hire to assess. An ongoing relationship with someone who has shipped production AI and will tell you the truth about it.
or €200/hr
A trusted technical perspective for your AI decisions: architecture reviews, vendor assessments, the questions your team can't yet answer. Five to eight hours a month on retainer with priority availability, or by the hour when you're not ready to commit monthly.
A more active role: sprint reviews, architecture decisions, mentoring your senior engineers, and representing the AI roadmap internally. For when you need an AI lead but not a full-time hire.
How payment works. Engagements are project-based, not hourly. Build work is milestone-gated: 30% to start, 40% at mid-point delivery, 30% on completion.
The diagnostic guarantee. If an AI Roadmap or Prompt System Audit delivers something you genuinely can't act on, you get your money back. The success criterion is binary and written into the contract before any work starts.