Work with me
I take AI prototypes to reliable production systems, hands-on. Engagements are remote and B2B, for US and UK companies.
Engagements
AI Readiness Review
Fixed scope · 2 weeks
A review of your AI prototype or plan: code, data, infrastructure and team. I identify what stands between it and production, including reliability, evaluation, cost, security and integration with your existing systems.
Ask about AI Readiness ReviewYou get
- Architecture assessment
- Prioritized risk list
- Production roadmap with estimates
Prototype to Production
Fixed scope · 4–8 weeks
Hardening an existing LLM or agent system so it holds up under real traffic: evaluation harness, guardrails and structured outputs, observability, cost controls, multi-model fallback and deployment.
Ask about Prototype to ProductionYou get
- Production-ready system
- Evaluation suite
- Runbook and handover
Agent Build & Embedded Architect
Ongoing · monthly
Designing and building autonomous agents and workflow automation with your team, or acting as a fractional principal architect for your AI platform.
Ask about Agent Build & Embedded ArchitectYou get
- Hands-on delivery
- Architecture decisions
- Code review and mentoring
How it works
- 1.Email a few lines about what you are building and where it is stuck.
- 2.We have a short call, then I send a written proposal with scope, timeline and price.
- 3.Work starts with weekly written updates and demos.
Questions
- Do you work with US and UK companies remotely?
- Yes. All engagements are remote and B2B, and I work from Europe. I have been consulting for US companies since 2014.
- What does an AI Readiness Review include?
- A two-week, fixed-scope review of your AI prototype or plan: code, data, infrastructure and team. You get an architecture assessment, a prioritized risk list and a production roadmap with estimates.
- How long does it take to get an AI prototype into production?
- A Prototype to Production engagement is fixed scope and takes 4 to 8 weeks. It covers an evaluation harness, guardrails and structured outputs, observability, cost controls, multi-model fallback and deployment, and ends with a runbook and handover.
- Can you join our team long term?
- Yes. The Agent Build & Embedded Architect engagement is ongoing and billed monthly: I design and build autonomous agents and workflow automation with your team, or act as a fractional principal architect for your AI platform.
- Which models and stacks do you work with?
- Claude, GPT and Gemini, often together behind one orchestration layer with multi-model fallback, in Python and TypeScript. I have also built custom PyTorch models for image, video and voice processing.
- How do engagements start, and how is pricing set?
- Email a few lines about what you are building and where it is stuck. We have a short call, then I send a written proposal with scope, timeline and price.