What you need to know:

• Bupa’s data strategy has shifted from hindsight to foresight. Under chief digital officer Ed Falconer, the health insurer has moved beyond predictive analytics and dashboards to build a real-time, AI-powered decision engine focused on prevention, not just response.
• The goal is to shape – not just observe – health outcomes. Bupa’s “digital health twin” model is designed to create real-time, data-rich profiles of customers that anticipate and act on health risks before they emerge.
• AI powers autonomous, personalised interventions embedded across the customer journey – triggering communications, allocating clinical resources, and initiating care pathways.
• After hitting a wall two years ago, Bupa streamlined teams, flattened structures, and focused on high-impact goals – delivering at 8x the speed and activating hundreds of AI use cases within a year.
• For its data teams, embedded, outcome-led collaboration allows for rapid-fire problem solving, ditching the escalation chain for on-site execution.
• As AI takes agency, questions about accountability move from theory to practice. Bupa’s answer: tie every decision back to purpose and patient impact.
• Falconer’s north star is clear – Bupa isn’t chasing better dashboards – it’s building systems that prevent illness, drive intervention, and redefine what it means to be data-driven in healthcare.

Bupa builds digital immune system

In a sector awash in dashboards and post-mortem analytics, Bupa has flipped the script from insight to intervention. Armed with a patient “digital health twin” and real-time AI engines, the insurer isn’t just predicting risk – it’s acting on it before symptoms show, at scale, and in real-time.

To succeed, the data team is hardwiring AI into the decision layer of healthcare as they look to develop autonomous systems that surface risk and trigger actions – without waiting for human approval. It also presages new risks for marketers, technologists and governance teams alike: when AI starts making the calls, who owns the consequences?

Profound shift

Speaking at a recent Databricks customer event in Sydney, Bupa chief digital officer, Ed Falconer was clear on one key point - it's a story about people and purpose, but it is also demonstrates a profound shift in how artificial intelligence is being used—not to explain the past, but to shape future patient outcomes. At the heart of the shift is a deceptively simple ambition: Use data to help people live longer, healthier, and happier lives. It wants to do this not just via better analytics, but through better decisions.

Welcome to the era of the preventative decision engine.

Real-time decisioning is not a new concept for Bupa. As Mi3 reported in 2023, Bupa initially found a fast path to value by deploying an AI-powered Pega decisioning engine that in just six months delivered strong incremental revenue growth, cost savings of circa $1m, a 50 per cent uplift in efficiencies and a 30 per cent improvement in customer responsiveness.

Eyes forward

Like many organisations, Bupa sees itself as a data-driven organisation, but it insights were often viewed through the rearview mirror. Churn models, attribution funnels, forecasting, while handy, were not transformative. What Falconer and his team have set out to build is something different: A platform that not only sees risk but initiates the next-best action. And these actions need to be autonomous, personalised, and embedded deep within the customer journey, he suggested.

"What we're really talking about is how we can deliver preventative, personalised health care to every one of our customers – and bring to life experiences that truly enable that care to be delivered, both in physical and critical environments when it comes to preventative care."

It’s all anchored in a concept Falconer called the “digital health twin". It's an idea where every customer is represented by a real-time, data-rich model – one that pulls together clinical history, wearable data, social demographics and policy interactions. Not just a record, it's a predictor and a decision-maker.

Per Falconer: "We can also start predicting what you’re likely to suffer from down the track, which leads us to develop preventative health care programs. For us, the key words are predictive, preventative, and personalised. The work we do in the data environment supports both customers and patients – they’re not always the same person or defined the same way – as well as clinicians and providers. If we do that well, we’re creating more time for better experiences and more meaningful conversations, which ultimately lead to better outcomes for our customers."

The success of the program hinges on significant culture change. "We have created a culture where data, people, and the craft of data science, data engineering, and data capability have come to life in support of our purpose at Bupa," Falconer told attendees.

Grinding to a halt

Two years ago, Bupa risked hitting the wall in its shift towards a more customer-centric operating system. "We’re obviously driving toward what everyone here would recognise as a customer-centric outcome. But the world has been completely challenged in terms of customer centricity because we have so many data silos. This is an issue many people in the room can relate to – data sitting in different pockets, across many systems. How do you bring all that together?"

Falconer’s diagnosis was that this represented classic enterprise drag: Too many touchpoints, too much noise, too little speed. So the company rewired itself – fast.

Team structures were flattened. Decision-making bottlenecks were dismantled. Falconer scrapped the kitchen sink approach in favour of focusing on a few high-impact goals. “We had to commit to fewer things and do them better,” he said. 

After 12 months, the team was starting to realise the benefits of the transformation, moving faster and with a huge volume of data migrated across into the new platform. "There is now a critical mass of data that the user base could access – what we call the critical data needed to run the business."

As a people leader, he said, "I saw our team become more engaged in their work – not because they weren’t aligned with our purpose before, but because they had been struggling to bring their craft to life. Now, they’re able to do their jobs and use their skills to actually extract value for the business, and as a result, their engagement levels rose significantly."

Autonomy became a defining principle. Objectives and key results (OKRs) gave teams structure without micromanagement. Partnerships – with internal teams, across squads, and with vendors like Databricks – were no longer transactional. They were embedded, outcome-focused, and execution-led, he said. 

The results were dramatic.

One year in, "We're going eight times faster", Falconer said.  More importantly, critical mass has been reached and the data platform is no longer a work in progress. That in turn, allowed for hundreds of use cases to come online – from retail cross-sell engines to real-time health risk stratification. 

"We have active use cases that are extracting and using data in meaningful ways. The more public-facing part of this is updating the predictive models we use in our retail environment – cross-selling, up-selling, those sorts of things – and making them much more real-time, accurate, and performance-based. We’re also bringing health models into the picture so we can better really understand the risk in the future, and therefore being able to put together preventative care."

Furthermore, the organisation is using Gen AI and other techniques to summarise huge amounts of data: "That helps us focus on the key features the data is telling us, and then recommend actions to take in preventative care."

According to Falconer, preventative care and preventative decision engines are being built so that not only are insights being generated, but so Bupa can rapidly deploy real actions and outcomes for the customer.

"We’re now at the point where we’re ready to scale with all the solutions and data we’re bringing in," he said. "Through that, we are really able to live up to our purpose –enabling our data scientists, data engineers, our data community, and the business as a whole to ensure data is helping our customers live longer, healthier, and happier lives."

Which is a little more meaningful than adding another dashboard.