Data and technology leaders can aim much higher with AI, and raise their own standing in the business as they do it. Here is why I think now is the moment to be bold.
For all the talk of AI overhauling the world of work, I believe most companies are still thinking too small.
Leaders need to look beyond AI productivity tools such as Copilot and Claude. These tools help people work a little faster, but when every competitor has the same technology and the same efficiency gains, they do very little to set a business apart. Many organisations also end up with fragmented tools that live in individual workflows rather than in company-wide systems.
There’s just so much AI noise that it’s really easy to focus on small, relatively well-trodden technology. Organisations are focusing on tools because they know what the output is and they can see how many people are using them. But that isn’t how you measure success.
If you can limit the conversation to tools, adoption and productivity, you can become braver about the fundamental changes AI can make. If you think bigger about how an organisation could serve its customers in the future, the impact is much greater than simply focusing on productivity for your employees.
Looking beyond productivity
At FOIL, business value is always our starting point and our end goal. We sometimes won’t take on projects that don’t have an obvious business value. If it’s just about being more productive, we don’t think it’s worth that company investing heavily in building the technology. They should just buy off the shelf.
What motivates us are the opportunities where AI can transform how an organisation operates and serves its customers. We would rather build lasting capability than sell consultant days. We build the technology, data foundations and internal capability that give an organisation real control of its AI, and we keep that technology evolving as its ambitions grow. We aim to get from experimentation to production value in cycles of just 12 weeks.
What our work with Welsh Water taught us
The clearest lesson I have learned about this came from our work with Welsh Water. Its Operational Response and Control team manages incidents across a 27,000-kilometre network, from burst mains to supply interruptions. Many of the pressure alarms it receives are nuisance alarms, and telling them apart from genuine incidents takes years of experience that lives in people’s heads rather than in any system.
Together we built an agentic system that triages those alarms in real time, assesses the potential impact, draws on operational knowledge and recommends a response to the human analysts. The models were not the hard part; understanding and mapping the context was. Once the context was in place, the impact could be immediately identified:
• 60 to 70% of nuisance alarms handled automatically
• 8,760 hours of manual triage work saved each year
• Faster incident identification
• Hundreds of thousands of pounds in projected regulatory penalties avoided
What made the difference is what we call digital threads, coined by our CTO, Simon Turner. Rather than only crawling numerical data, the system parses unstructured knowledge and maps cause and effect across related events, so the agents understand not only what is happening, but why it matters. That is why many say, without context, your AI is worthless.
The project went on to win Agentic AI Solution of the Year at the 2026 British Data Awards. For me, though, the real story is the knowledge work behind it. So much of the knowledge people rely on is tacit. It exists in conversation or in the experience people carry around in their heads. So how do you capture that and store it in a way that makes it reusable for AI? That’s something FOIL has been doing for years, from Welsh Water to Amey’s agentic Knowledge Bank. We treat knowledge capture as part of the build rather than an afterthought.
This ability to turn institutional wisdom into working technology is what led us to a concept I believe in deeply: the autonomic business.
What is an autonomic business?
An autonomic business behaves like a living system. Your nervous system raises your heart rate when you climb the stairs without you asking it to. In the same way, an autonomic business senses conditions in real time, makes or recommends decisions within agreed guardrails, adapts based on feedback and learns to improve. Intelligence is embedded in how the organisation runs, rather than bolted on.
None of this means handing your company over to AI. People stay in the loop, setting the direction and managing the agents, while AI takes on the repetitive, everyday work. Find out more about the autonomic business.
Reimagining, not replicating
If this sounds futuristic, you are not alone. A few years ago, even some of our most forward-thinking clients struggled to believe it was possible, and I found the concept hard to connect with at first too. Today, agents are everywhere and technology is no longer the biggest barrier. Company culture is. Transforming a business is as much about people and the operating model as about technology.
That is why transformation should not happen all at once. You take one autonomic capability, build it at MVP level as you would any other data asset, and then develop it and develop it and develop it. Over time, people see the impact, understand the opportunity and start to invest in the vision.
The hard work of change still has to happen. An autonomic business isn’t about recreating the processes that happen today. It’s about looking at those processes and asking: why the hell are we doing that? We should actually rethink it.
Giving data leaders a seat at the table
For years, businesses have treated data and technology teams as an afterthought. Someone asks a question, the data experts try to answer it, and then they take the blame when fragmented systems or poor-quality data get in the way.
AI can change that, but first organisations need to raise the role data leaders play. Data leaders should be in executive meetings, helping make decisions about where the business is going and providing insights the business has never seen before. AI is a business topic, not just a technology or data topic.
Many businesses are not ready for that yet. We all know it’s hard to keep the excitement high when you’re making large changes in data capability, because you can wow the board with AI tools but then you need to build the foundations – and that does take time. If an organisation does not lift its data capability alongside its business strategy, the board will tire of the work very quickly.
This is where strategy earns its place, even though many data leaders see it as almost a dirty word and picture a 100-slide deck with no substance behind it. We turn a data problem into a narrative that engages the whole board, so CDOs can define the business change they want and secure lasting buy-in for it.
Our work with an insurance company shows what this looks like. AI may remove the need for brokers in the long term, so together we faced the existential questions: how might the business change, what would that mean for its customers, and what kind of company would it need to become? That work quickly gave the Chief Data and Strategy Officer a stronger voice with the board. We built that narrative with her, and now we’re developing the roadmap to follow.
The vision should be for the business, and it should be underpinned by data and AI. My colleague Mish Naik explores this further in Traditional data and AI strategy is dead. But what replaces it?.
Blink and you have 17 agents
Skipping the strategic work also creates shadow AI. For years, organisations have battled the curse of multiple spreadsheets doing the same job in different corners of the business, with major decisions resting on them. The same thing is now happening with agents. Everyone can build one. You blink, and suddenly the whole company has 17 productivity agents doing the same work. People are using them privately, they’re not really sharing them, and there’s no process around any of it.
My advice is to start with one AI agent that everyone can access. Better still, plan for a platform that orchestrates all of your future agents so they work together, with shared memory and governance, as connected entities rather than isolated tools. Companies need to think strategically, rather than focus on isolated use cases. Otherwise, you’re just going to end up with the same Excel drama and chaos all over again.
A challenge to the industry
Most organisations want a different story but do not always know how to write it. We recently received a request for proposal with a prescriptive brief focused on largely out of date technical foundations and a bit of AI. We presented the autonomic business concept instead, encouraged the organisation to think differently, and won the tender.
It proves that organisations want to think differently. But unless they’re challenged and shown what the business could actually look like, they become stuck in the weeds. Data and technology leaders have a once-in-a-generation opportunity to define how companies will look in ten years.
My call to action is simple: be bold, but have a journey to get there. If you want to explore what that journey could look like for your organisation, get in touch.
A version of this article first appeared in Orbition Group’s Driven by Data magazine.