Enstar Group operates in one of the most complex and highly regulated segments of financial services. The global re/insurance company specializes in retrospective solutions and specialist underwriting, helping insurers and reinsurers manage complex risk portfolios, legacy liabilities, and capital requirements across international markets. Operating across major insurance hubs worldwide, the organization relies heavily on accurate, governed, and scalable data to support decision-making and manage risk effectively.

For David Tuppen, the organization’s Chief Data Officer, modernization is no longer simply about replacing old technology stacks or introducing new tools. It requires building foundations that business teams can trust and use.

In the first part of a two-part series, Tuppen speaks with Robert Lutton, Vice President at Sandhill Consultants, about the true mandate of the CDO function, what modernization should really mean, and why AI is exposing weaknesses in enterprise data foundations.

For Tuppen, success in the CDO role is defined by whether the business actually uses what the data office builds. “Adoption by the business is the key winning factor, and that’s a mandate that any data ops should have,” he says.

Building foundations the business can use

Tuppen emphasizes that a data foundation is not merely a technical construct. While organizations often focus on building a data platform, architecture, data lake, or warehouse, he argues that the real foundation begins with the right team and capability.

According to him, organizations need a team that can support what is being built, followed by the architecture, design, and platform. Once the platform exists, the next priority is managing it through governance and data protection. Only then can organizations build meaningful insights, predictive analytics, and embedded AI.

“The mandate that CDOs and data offices should adopt is integrating data within business teams,” Tuppen states.

Modernization is simplification

Tuppen notes that modernization is often misunderstood. Many organizations view it as moving from traditional warehouses to lakes, lakehouses, or next-generation architectures. But he believes the real issue is much simpler.

Modernization, in his view, is about reducing complexity, removing manual work, and creating a cleaner path for data to move through the organization.

That means looking closely at how data enters the platform, where it is stored, how it is modeled, and how it is distributed to business units and consuming technologies. Tuppen points especially to the manual processes and Excel-based workarounds that still exist inside many enterprises.

“It’s simplifying the manual intervention, moving towards automation,” he says. “That’s what we should be focusing on when we talk about maturing the estate or modernization.”

Governance cannot be an afterthought

In regulated industries, modernization cannot come at the expense of control. Tuppen says governance and protection must be built from the beginning, not added later as a corrective measure.

He connects this directly to the idea of strong data foundations. A foundation includes not only platforms and architecture but also the management of data through governance, privacy, and protection.

“You don’t catch up with data governance. Data governance isn’t a tech debt that you want in two to three years after building a platform,” he says. “You build privacy and protection by design. You build governance up front.”

For regulated environments, this is not optional. It is the way modernization can move forward without creating risk.

AI exposes weak data foundations

The rapid adoption of AI has increased pressure on organizations to modernize, but Tuppen does not believe AI has fundamentally changed the goal. Instead, he says it has made existing weaknesses more visible.

“AI has exposed weak data foundations,” he says.

He explains that large language models (LLMs) depend heavily on the quality and context of the data they receive. Poor data leads to poor results, and vague prompts without enough context fail to deliver what users expect.

“If you throw in poor data, you get poor results,” he says.

For Tuppen, the lesson is clear: Without clean, governed, and controlled data, AI remains limited. “Clean up your data, have that data foundation in place, and then the AI will come into its own.”

Adoption comes from business ownership

As the conversation turns to enterprise-wide transformation, Tuppen identifies business ownership as the real driver of adoption.

He cautions against operating models where the data office controls every aspect of data accessibility or simply supplies data to business units. Instead, business teams must be involved in creating and using their own data products and insights.

With AI increasingly embedded into enterprise technologies, Tuppen believes business teams will naturally use it once the right data foundation is in place.

In conclusion, he says, “Once the data foundation is in place, business teams will apply AI to their own data. The key is giving them the capability and ownership to build solutions around it. Real adoption happens when the business owns both the data and the outcomes. Give teams access to their data, and they will use it.”

CDO Magazine appreciates David Tuppen for sharing his insights with our global community.