Founder & CEO at Digital Twin AI | Lead Engineer | Microsoft Certified Professional | IT Specialist | Business Strategist | Project Manager | Team /Startup Leader
The first version of what would become one of the UK’s most distinctive AI systems did not run on a factory floor. It ran on the imagination of two six-year-old boys. Brian Stack’s twin sons were obsessed with dinosaurs, and Stack, who lives with dyslexia and ADHD himself, understood from his own experience that the standard routes to knowledge are not always routes at all. He knew firsthand how it feels when information is locked behind formats that simply do not fit the way your mind works. So this father and lead engineer, who spent his days maintaining complex production lines, spent four and a half years building his boys something different: T Rexxy, an AI dinosaur that could talk, explain, and teach in a way that met them where they were. In T Rexxy’s own words, “Raww! It’s a big wide world, let’s go on an adventure together!”
That single sentence contains the entire philosophy that now drives a serious industrial AI company, one positioned to become Britain’s sovereign AI provider. Knowledge should not be locked behind formats that exclude the people who need it most, whether those people are young boys discovering the world or skilled engineers standing in front of a stopped production line at two in the morning. And in Stack’s view, the intelligence that delivers that knowledge should belong to Britain, not be rented from servers overseas.
The Engineer Who Built What He Needed
Brian Stack is the Founder and Chief Executive Officer of Digital Twin AI, a UK-based company building what is set to become Britain’s sovereign AI provider, giving industrial machinery an intelligent, conversational voice that runs entirely on British soil. He is one of the rare founders whose product exists because he could not find it when he needed it most, and because he had spent a lifetime experiencing the problem from the inside.
Twenty Years of Hands-On Solutions
Stack did not arrive at artificial intelligence through theory. He arrived through twenty years of getting his hands dirty with complex systems under pressure.
His career began at Global Integrity, designing bespoke IT and surveillance solutions for international clients and responding to cybersecurity incidents across Europe. The work required building custom hardware from the ground up and securing compromised networks when the stakes were at their highest. From there, he moved to advanced UAV systems at Clogworks Technologies, where he designed what would become the world’s first 5G drone platform and built disaster relief communication infrastructure for the United Nations.
It was his role as Lead Engineer at Richard Carter Ltd, a manufacturing company with heritage stretching back to 1740, that brought the central problem into sharp focus. Stack was responsible for every engineering aspect of the company’s operations, and he systematically brought critical capabilities in-house that had been outsourced for years.
Copy lathe models that previously cost between £5,000 and £20,000 per unit, with lead times measured in weeks, were redesigned and 3D printed in-house for approximately £75, with lead times dropping to hours. When a plastics molding supplier suddenly closed, Stack and his team spent two intensive weeks rebuilding the entire department on-site. “My objective was clear: to bring critical engineering capabilities in-house and eliminate reliance on expensive third-party services,” he explains. The same instinct, bringing critical capability home rather than depending on outside suppliers, would later define his entire approach to AI.
But the deeper insight came from watching skilled engineers lose hours hunting through dense manuals while production lines sat stopped. “The most valuable asset in any factory is not the machine,” Stack says. “It is the time of the skilled engineer needed to fix it.” He had already spent four and a half years building an AI system that could make knowledge accessible to children who learn differently, drawing on his own understanding of what that takes. The connection to the factory floor was inevitable.
Giving Every Machine a Voice That Engineers Can Trust
Digital Twin AI’s core proposition is precise: transform technical documentation into a conversational expert system that any engineer can query under real production pressure. The results at Richard Carter Ltd demonstrate what that means in practice. An electrical diagnosis that previously required eight to nine hours now takes nine minutes.
The process behind that result sets Digital Twin AI apart from most industrial AI implementations. Stack begins every client engagement with a pre-assessment that many competitors skip entirely. He audits the client’s foundational processes before any AI is deployed because, as he puts it, “you never know what you are going to get unless you find the manual.”
This is where Stack’s engineering integrity becomes most apparent. He will tell a prospective client directly that they may not need his product yet. If spare parts inventory is reactive, if the flow of information between engineers is broken, if core operational workflows are flawed, then reducing diagnostic time will not matter. “Everyone is eager to sell, but not honest enough to say: if you fix this, you might not even need AI for that specific problem,” he says.
Once foundational processes are sound, the Digital Twin AI system ingests everything: dense manuals, schematics, standard operating procedures, historical maintenance records, and the tacit knowledge captured from experienced engineers. Each machine receives its own AI agent, trained exclusively on that machine’s data. Critically, every answer includes direct references to source documentation, with page numbers and section titles, so engineers can verify recommendations under pressure.
“Just like Captain Picard asking the ship’s computer for a status report, our system acts as the living, breathing embodiment of the machine,” Stack explains. “Trust is not assumed. It is built into the architecture.”
Britain’s Sovereign AI Provider
This is the heart of what Digital Twin AI is set to become. The system operates entirely offline, with no reliance on foreign cloud infrastructure, making it the UK’s only fully sovereign AI system of its kind. Every model, every piece of data, and every answer stays within British control, never leaving the country and never depending on a foreign provider. For manufacturers handling sensitive industrial data, and for a nation increasingly aware of the risks of depending on overseas technology, that is not a technical footnote. It is the whole point.
Stack is unambiguous about the ambition. “Britain should not have to rent its intelligence from abroad,” he says. “We are building the sovereign AI capability this country needs, owned here, run here, and answerable here.” It is a vision that places Digital Twin AI at the front of the national conversation about technological independence, at exactly the moment that conversation has become urgent.
That distinction has already drawn attention. At an invite-only AI startup conference at Google’s London headquarters, Stack was one of sixty founders and the only one with a 100% UK sovereign large language model. “I thought I would be the underdog,” he says. “Instead, I had CTOs and VCs telling me that sovereignty and real factory traction were exactly what they had been looking for.” The validation reinforced his conviction that practical results, delivered on sovereign foundations, matter more than theoretical capability.
The company’s sister platform, Aigenthive, extends the same sovereign AI architecture beyond factories, creating expert agents trained exclusively on curated educational materials. “Our AI will not tell you when Shania Twain released her last album,” Stack says with a smile. “But on the field it is trained for, it will be the leading authority.” Together, the two platforms point to a single ambition: a sovereign AI capability, built in Britain, that can serve the factory floor and the classroom alike.
The Problem He Solved Twice
Digital Twin AI is now three months from full rollout into Phase 2 of its partnership with Richard Carter Ltd. Eight days into 2026, Stack was already at the BETT Show on the UK Government’s pavilion, demonstrating how the same system that reduces electrical diagnosis from nine hours to nine minutes can make knowledge accessible to every engineer who learns differently.
T Rexxy was never just a toy. It was proof that knowledge, properly structured and intelligently delivered, can reach anyone, regardless of how they learn or how much time they have. Brian Stack spent four and a half years building that proof for two boys who needed it, guided by his own experience of learning differently. He has spent the years since building it for every engineer standing in front of a machine that has gone quiet, and building it as something Britain can call its own.
The factory floor and the six-year-old’s bedroom turned out to be the same problem all along. So did the answer: intelligence that belongs to the people who depend on it.
Brian Stack is the Founder and Chief Executive Officer of Digital Twin AI, based in New Mill, England. Digital Twin AI is building Britain’s sovereign AI provider: offline, UK-owned AI systems that give industrial machinery a conversational expert voice, serving manufacturers who cannot afford downtime, unreliable answers, or dependence on foreign technology. To connect with Brian or learn more, visit www.digitaltwinai.co.uk or www.linkedin.com/in/brian-stack-dtai.
