I Build AI Systems That Know
What Only Your Company Knows.
There are organisations sitting on data that took decades to accumulate — mineral deposits, clinical knowledge, industrial inventories, proprietary research. Generic AI cannot touch it. Most technical teams cannot build on it properly. A very small number of projects earn my full attention. If yours might be one of them, it's worth a conversation.
What Is Sovereign AI
Your data is already
extraordinarily valuable.
Most companies just can't ask it questions.
ChatGPT is impressive. It also knows nothing about your mineral deposits, your compliance frameworks, your industrial property database, or the 20 years of clinical expertise sitting in your knowledge base.
The organisations winning with AI aren't using public models for competitive advantage. They're building private intelligence layers on top of data their competitors don't have.
A Sovereign AI system sits on top of your proprietary data and turns it into a competitive weapon — one you own outright, that runs on your rules, and that no vendor can switch off, price-hike, or deprecate.
That's the gap. And most organisations don't have the technical infrastructure — or the time — to close it themselves.
The Sovereign AI Stack
Live Systems
Three systems. Running now.
Not case studies. Not prototypes. Production intelligence systems interrogating real proprietary datasets — built and deployed in weeks.
Doctorate-Level Knowledge Base
A PhD-level cybersecurity thesis — years of clinical research, frameworks, and methodology — restructured as a Sovereign AI coaching system. Users interrogate it through an iPhone app. The expertise scales without the expert needing to be in the room.
IndiFind — European Industrial Property Intelligence
A pan-European database of industrial property assets, overlaid with an AI that lets investors, developers, and brokers ask complex questions in plain English. What took an analyst two hours now takes two minutes. The data advantage stays inside the platform.
Global Minerals Intelligence System
A living database of global mineral deposits, supply chains, and asset valuations — built for investors who need to move fast in a market where data advantage is everything. The AI doesn't just surface data. It reasons over it.
How It Works
Four stages. Typically 4–8 weeks.
No prototype theatre. No six-month strategy documents. Every engagement ends with a production system you own.
Data Architecture Review
We map your proprietary data — structure, gaps, quality, and intelligence potential. I tell you honestly what an AI can reliably do with it, and what it can't.
Intelligence Layer Build
Data pipeline, retrieval architecture, reasoning layer, and interface — web, mobile, or API. Built with production-grade infrastructure from day one.
Deploy & Refine
Your system goes live. I monitor accuracy, tune the model behaviour, and document everything so your team can maintain and expand it.
Handover — You Own It
Complete documentation, clean codebase, zero vendor dependency. No subscription on the other side. No renewal. No lock-in. It's yours.
The Standard
Not every dataset warrants a Sovereign AI system.
Yours might.
The projects I take on
What won't clear the bar
Why Me
19 years building systems that run the world's data.
19 years building the systems that manage critical data at scale — at Thomson Reuters, Woolworths, the Australian Government, and organisations where the technical decisions have real commercial consequences. I've implemented the enterprise platforms that most companies rely on, which means I understand both what they're worth and exactly where they fall short.
I work at the intersection of deep data architecture and AI-accelerated delivery — a combination that doesn't exist inside agencies, can't be hired full-time for under £150k, and takes years to develop. The result is a kind of engagement most organisations have never had access to: one person who can design the intelligence architecture, build the system to production standard, and hand it over — without committees, without contractors, without delays.
TOGAF 9.1 certified. Former Top Secret clearance. 219 projects across 19 years. The track record is not theoretical.
What decision-makers say.
Exceptional professional — expertise in enterprise architecture, Salesforce, PMO leadership. Remarkable ability to simplify complex projects. Transformed our PMO.
Exemplary platform knowledge with dynamic, thought-provoking approaches. Exactly what large enterprises need early in their Salesforce journey.
Highly professional architect — strong in technical and business engagement. Breaks down complexity, builds relationships.
Knowledgeable and well-connected professional — scoped and addressed company needs accurately and timeously.
Tight-knit team delivered on tight timelines under pressure. Focused on getting right things done right — project shipped, users happy.
I replaced a £140k/year Salesforce stack in 6 weeks. Strategic, precise, and completely focused on what actually ships — not what sounds impressive in a proposal.
Services
Two ways to work together.
Every engagement ends with a production system you own outright. No subscriptions on the other side.
Proprietary Intelligence System
For organisations with specialist data that warrants a bespoke intelligence layer. Architecture, build, deployment, and handover — delivered as a single engagement by one person, with no agency overhead and no handoff delays. You own the system outright on day one.
- Data architecture review and intelligence potential assessment
- Full data pipeline — ingestion, cleaning, indexing
- Reasoning layer and retrieval architecture
- Interface — web application, mobile app, or API
- Deployment, accuracy monitoring, and tuning
- Complete documentation. Full IP transfer. No ongoing dependency.
SaaS Detox Sprint
For established businesses where SaaS costs have grown beyond what the value justifies. A structured audit and custom build — replacing the platforms you've outgrown with a system built exactly for how you operate. Delivered in 4–6 weeks. No subscriptions on the other side.
- Complete SaaS stack audit — spend vs. actual usage
- Custom replacement build (Next.js, TypeScript, Supabase)
- Data migration and team onboarding
- Full cost analysis — 3-year ROI documented
- Technical documentation. You own it outright.
Selective Enquiry
Most enquiries don't proceed.
The right ones do.
I work with a very small number of organisations at any time. Engagements are accepted based on the quality of the data, the clarity of the commercial case, and whether the project is genuinely interesting to build. If you have proprietary data with real strategic value and the ambition to turn it into a permanent competitive advantage — it's worth finding out whether this is the right moment to talk.
There is no sales process here. Either the project clears the bar, or it doesn't. The conversation will tell us both quickly.
30 minutes.
I'll tell you honestly whether it's a fit.