In an era where nuclear energy is being called upon to meet global decarbonisation and energy security goals, the workforce behind it is being stretched thin. Between regulatory pressure, aging infrastructure and increasingly complex documentation requirements, nuclear professionals spend a growing portion of their day on paperwork – not engineering.
Nuclearn is aiming to change that by harnessing the power of AI. Co-founded by Brad Fox (CEO) and Jerrold Vincent (CFO), Nuclearn was created because the pair were exhausted by inefficiency within the nuclear industry, and thus decided to create a platform providing private, secure AI solutions designed to accelerate documentation, streamline compliance and increase engineering capacity.
Go deeper with GlobalData
Discover B2B Marketing That Performs
Combine business intelligence and editorial excellence to reach engaged professionals across 36 leading media platforms.
The platform, which they say is grounded in plant reality rather than Silicon Valley theory, has already been deployed across more than 50 facilities in the US, Canada and the UK. David Appleyard speaks with Fox and Vincent about why it matters now – and how AI can make nuclear power not just more efficient but more effective.
David Appleyard (DA): Let’s start with the big picture – what problem is Nuclearn solving in nuclear today?
Brad Fox (BF): It really starts with the people. Nuclear was never supposed to be this bloated from a staffing standpoint. Go back to the 1950s and 1960s, those early plant designers were imagining that these facilities could be run with 300 or 400 people. But today, many sites are staffing 700–800 or more per reactor.
The problem isn’t safety or operations. The bloat is in documentation, compliance, procedure writing and administrative burden. That is the hidden cost of nuclear power. You are paying highly trained engineers and operators to do manual, repetitive tasks that are necessary but slow and frustrating.
We founded Nuclearn because we were living through that friction. We believed there had to be a better way, and AI, properly applied, offered the answer.
US Tariffs are shifting - will you react or anticipate?
Don’t let policy changes catch you off guard. Stay proactive with real-time data and expert analysis.
By GlobalDataDA: Did this come from your personal experience inside the plants?
Jerrold Vincent (JV): Absolutely. Brad and I both spent years in nuclear operations. We were on the floor. We wrote the reports, sat through the reviews, got the same procedure kicked back five times because a citation was one character off.
It became obvious that most of the work we were doing wasn’t technical, it was matching patterns. It was compliance logic. That’s where AI shines.
We started small: document automation, CAP [corrective action programme] screening, repetitive risk assessments – but as we built trust and capability, the demand grew. Now, Nuclearn’s platform supports use cases across maintenance, safety, engineering and regulatory functions.
DA: What does the platform actually do – and how is it deployed?
BF: At its core, Nuclearn is a private AI ecosystem that connects natural language processing with plant-specific data, processes and security models. Our customers use it to write and validate procedures, analyse safety observations, classify work for capital versus expense, perform document-based research across licensing and QA archives, assist with outage prep and planning, and more.
It is available as a private, on-premise system or a secure hosted model, built with compliance in mind. Everything is permission-based and auditable. The data stays private and every customer environment is isolated.
DA: How does Nuclearn ensure it is “nuclear-grade” when it comes to AI?
JV: What we have built isn’t a generic AI model. It is a platform trained on nuclear data, regulatory structures, procedural formats and real-world use cases. Our AI doesn’t hallucinate because it is grounded in context. If you are pulling up a licensing document or CAP history, the system isn’t guessing. It is searching verified documents in your environment and surfacing real excerpts.
We have also built in what we call ‘human-centered workflows’. That means the AI isn’t replacing the expert; it is partnering with them. You are always in control, verifying the output, editing as needed and reviewing before submission.
DA: Nuclearn pushes the concept of an AI marketplace for nuclear – what does that look like?
BF: That is where it gets exciting. Right now, every plant is solving its own problems in isolation. A utility in Georgia may have built a brilliant AI-based outage prep tool, but no one in Illinois knows about it. What we are building is a secure, nuclear-specific AI marketplace where utilities can share, buy or adopt pre-trained models, prompts and apps. It is a way to accelerate innovation across the industry while respecting IP [intellectual property] and data privacy.
The marketplace also supports community-driven learning. As more plants use similar tools, we can build in best practices and improve the model quality over time.
It is open innovation but within a secure, nuclear-compliant framework.
DA: Tell us about your work beyond the US.
BF: Right now, most of our deployments are US-based, but that is changing quickly. We have had strong traction in Canada with Canadian Nuclear Safety Commission-regulated sites, and we are starting to engage customers in the UK as well.
The beauty of our system is that it is customisable. Even if the base AI was initially trained on US regulations, our clients can input their own documents – regulatory guides, site manuals, licensing histories – and the AI can search and reference those sources. So even before we have fully trained an international model, the system is still delivering value.
JV: We are working on that now. One of our goals for the next year is to build a more comprehensive international model. That means incorporating UK Office for Nuclear Regulation standards, Canadian licensing terms and other country-specific guidance into the core AI engine.
We have already laid the foundation with our architecture. Now it is about scaling that intelligently.
DA: How do you address the perception that AI is risky in high-consequence environments like nuclear?
JV: It is a legitimate concern, and it is why we have built Nuclearn to support – not replace – human judgment. We always say: in nuclear power, the operator is the hero. The engineer is the expert. Our platform exists to support them, giving them back time, reducing cognitive load and letting them focus on the things that actually require human insight. Unlike many AI tools, ours is not a black box. Everything the AI produces is traceable. You can click and see where it pulled that citation from, what revision it used and how it matched the prompt.
BF: That is why we have gained trust quickly. Our early adopters didn’t just see a flashy demo, they got results. Fewer documentation errors. Faster turnarounds. Higher compliance confidence. When you show up with a secure, auditable, proven solution that saves people hours a week, the scepticism fades.
DA: Are there some real use examples that stand out?
BF: A few stand out. One team reduced their procedure prep time by 50% using our AI assistant. Another used our capital versus expense classifier to improve how work was coded, which had a material impact on their income statement.
We have seen safety analysts go from spending hours compiling observation trends to having a dashboard ready in minutes. Maintenance planners are using the system to streamline outage scope planning. It is widespread.
JV: Those are just the formal use cases. Informally, we hear things like: “It helps me get unstuck.” Or “It makes me feel more confident that I haven’t missed something.”
That is when you know the AI is doing its job – it is reducing friction and building trust.
DA: Where do you see the role of AI in nuclear heading over the next five years?
JV: We believe AI will become as essential as the control room. Not because it runs the plant, but because it frees the people who do. You will see AI embedded into safety analysis, licence renewal planning, outage coordination, even vendor audits. Not to replace experts but to make their expertise go further. That is critical, especially as we face workforce attrition and an aging knowledge base. More importantly, we are listening. Every feature we have built has come from a conversation with a utility, an operator or an engineer who said: “What if it could do this?” That is what keeps us essential – not just being smart, but being in service to the mission of nuclear.
BF: I think the bigger picture is: if we want to scale nuclear, we have to scale the human element too. We can’t just build more plants, we need to operate them with smarter processes, tighter workflows and better use of our people’s time.
AI done right is how we get there. As the nuclear sector contends with aging plants, workforce turnover and the need to scale new generation, the team at Nuclearn aims to set a new standard, not just for what AI can do but for how it should be done. If AI is to play a role in the future of energy, it will need to earn the trust of the most exacting industry on Earth.
