Most enterprises have modern data platforms, yet few see the AI value they expected. Gartner predicts that 60% of AI projects will be abandoned through 2026 due to a lack of AI-ready data, and McKinsey reports that nearly two-thirds of firms still fail to scale AI because of fragmented data and poor quality. The shift to data products is emerging as the answer: domain-owned, discoverable, trusted assets built for consumption, not reporting.

This panel brings together leaders who have made the transition from centralised, request-driven data models to federated, product-oriented organisations capable of supporting AI at enterprise scale. They’ll discuss what the operating-model change really required, how they overcame resistance, and what “good” looks like once data products become the backbone of AI delivery.

Nadia HussainBusiness Transformation DirectorMacMillan

Niranjan UmashankerGlobal Head of Data IntegrationL'Oreal

Apoorva KumarSenior Director - Strategy & TransformationPepsiCo

Keiran HollowayChief AI ArchitectRackspace Technology

Aaron KalvaniGlobal AI StrategistUnited Nations

AI adoption is widespread, but true enterprise-level impact remains rare. McKinsey’s 2025 survey shows that 88% of organisations use AI, yet only one-third have scaled it, with 39% still stuck in pilots. Despite IDC reporting an average 3.7x ROI on generative AI investments, BCG finds that 60% of companies see little or no material value, while the top 5% of AI leaders achieve 5x the revenue gains of their peers. The difference isn’t tooling, it’s leadership ownership, measurement discipline and the ability to turn pilots into repeatable production blueprints.

In this rapid-fire Lighthouse Showcase, three senior data and AI leaders each present one initiative that successfully moved from experimentation to scaled value, in just 5 slides and 10 minutes. They’ll share the metrics that mattered, the governance that unlocked funding, and the operating-model shifts that enabled production-grade AI.

A high-density, insight-rich session for leaders ready to move beyond pilots and build AI programmes that compound in value.

Tia CheangFormer Chief Data & AI OfficerInvestIndustrial

Rajen SaraswatiChief Data ArchitectHome Office

Ramy ErfanVP Business & Technology EnablementCiti Group

Chris ShireHead of Architecture & PlanningEngland Rugby

Aaron KalvaniGlobal AI StrategistUnited Nations

Deploying AI introduces new levels of complexity across enterprise technology environments. As organisations move from experimentation to production-scale AI, traditional monitoring and management approaches are struggling to keep pace with increasingly dynamic, distributed, and autonomous workloads.

This session explores why technology leaders must build observability directly into their AI roadmaps to ensure system reliability, maintain operational control, optimise performance, and manage cost at scale.

Attendees will gain practical insight into how observability is evolving to support AI-powered applications, helping organisations identify issues faster, improve user experiences, and confidently scale AI initiatives across the enterprise.

Russell BrownSenior Solutions ConsultantNew Relic

A GT3 car generates more data in a single qualifying lap than most teams can read before the next one starts. The driver has about 90 seconds and no second attempt. The engineer has a pit wall, a weather radar, a tyre model and a race that will not pause while he decides.

Jamie Falvey raced Team BRIT's BMW M4 GT3 in the 2026 GT Cup Championship, finishing the season as GT3 Champions. Matt Cherrington, director of TMC Race Engineering, engineered it. Between them they do, every race weekend, what most enterprises are still trying to build: turn a flood of telemetry into a decision that can be made in seconds, defended afterwards, and repeated next time.

In this fireside chat they take the room inside a race weekend: how a tenth of a second is found in a corner, where simulation ends and judgement begins, what they do when the forecast is wrong and the safety car is out, and when the right call is to ignore the data entirely.

It is not an enterprise case study. It is the same problem, at 150mph, with the consequences of being wrong arriving immediately.

Seakers, Jamie Falvey, GT3 Racing Driver, Team BRIT

Moderator, James Alliband, Chief Marketing Officer, Technology Leaders Club

Jamie FalveyGT3 Racing DriverTeam BRIT

Matt CherringtonRace EngineerTeam BRIT

James AllibandChief Marketing OfficerTechnology Leaders Club

An AI can give a convincing explanation for a wrong answer, and fake citations in published research are rising fast. In this session, Pascal argues that a confident AI isn’t a trustworthy one, and that organisations need to be able to prove why an AI output was safe to rely on. He sets out a practical eight-step framework for answers you can stand behind: ground them in good evidence, trace them to their source and keep logs, then verify, update, escalate and audit them. He then looks at the bigger shift, as employees, customers and their AI agents learn to work around corporate systems faster than policies can keep up. You’ll leave knowing how to govern the whole AI system, not just the model.

Pascal HetzscholdtSenior Director, AI Strategy & Content IntegrityWiley

Synopsis: Your winning AI formula isn't a product or a prescribed stack. It's the right combination of your data, technology, models, platforms and expertise for the outcomes you need. Rackspace Technology helps tech leaders make those choices, bring the pieces together and operationalise them at scale, so they can focus on where AI creates genuine business advantage.

Keiran HollowayChief AI ArchitectRackspace Technology

Description: AI success depends on a range of factors: scalable technology, reusable agents, effective model governance, business value and cultural change. But beneath all of these sits a more fundamental challenge: is your enterprise data ready for AI?

Without semantic context, rich metadata and trustworthy data, even the most ambitious AI strategy is built on fragile foundations.

The problem is that traditional data governance has often failed to keep pace. It remains disconnected from the analytics, AI and data engineering communities it needs to enable, burdening teams with requirements rather than making good governance easy to adopt.

In this talk, Kat will show how that can change. She will explore where innovation can remove friction, where governance needs to evolve, and where tempting shortcuts ultimately undermine AI readiness. The goal: a practical blueprint for accelerating enterprise data readiness for AI.

Kat HolmesData DirectorSiemens Energy

Most enterprises have a responsible AI policy. Far fewer can point to a decision it actually changed.

The gap between principle and practice is closing fast, not by choice, but by necessity. Gartner predicts 40% of enterprise applications will feature AI agents by end of 2026. Systems that act autonomously, often without a human at the moment of decision. PwC found that 60% of executives believe responsible AI boosts ROI and efficiency, yet nearly half say turning those principles into operational processes has been their biggest challenge. And McKinsey's research is unambiguous: organisations seeing real AI returns are three times more likely to have strong senior leadership ownership of AI governance. Responsible AI is not a constraint on performance, it is a precondition for it.

This fireside chat cuts through the framework debate and into the operational reality. How do you embed governance into the way engineers build and models are monitored? How do you maintain human oversight when agents are making thousands of decisions a minute? And how do you measure whether your responsible AI programme is actually working?

Devika MohanHead of Digital TransformationNHS England

Pankaj SabhlockHead of Data Products, AI & Digital TransformationUniversity of East London

Ian CarpenterHead of Data & AI GovernanceHeathrow

Aaron KalvaniGlobal AI StrategistUnited Nations

The 2025 AICDI Global Insights Report shows that less than a third of companies have a dedicated team for AI governance, and only 13% have policies ensuring human oversight, revealing a major leadership and operating-model gap.

Modern enterprises require a unified leadership blueprint where data, technology and security operate as one.

This fireside chat explores how the roles of CDO, CIO and CISO are converging as AI becomes mission-critical infrastructure. We examine how leading organisations are redesigning operating models, governance forums and accountability structures to balance innovation, security, resilience and compliance.

Zaman HussainChief Data OfficerGreat Ormond Street Hospital

Tim GrievesonChief Security & Risk OfficerCloudPay

Alistair CampbellGroup IT Director (CIO)Kindertons

Aaron KalvaniGlobal AI StrategistUnited Nations

Ban AI and people just stop telling you they’re using it. Rich Merrett, Head of AI at Bellrock and award-winning AI leader, shares a practical blueprint for empowering a whole workforce, and for governing AI so it can scale safely.

Richard MerrettHead of AIBellrock

McKinsey’s 2025 survey shows that half of high-performing organisations are redesigning workflows to integrate AI, and workforce transformation is now a top CEO priority. Reserach also highlights that only 18% of organisations report high levels of data literacy across all business roles.

This session explores how organisations are addressing the widening skills gap across data engineering, ML engineering, AI product management, governance and literacy. Our speaker will share practical strategies for reskilling, redesigning roles, embedding AI into daily workflows and building a culture that embraces AI, not fears it.

Richard MerrettHead of AIBellrock

Jesus Requena CarrionHead of Data Science & Engineering ReseachQueen Mary University of London

Naman GoyalAVP - AI Product & StrategyPL Capital India

Aaron KalvaniGlobal AI StrategistUnited Nations

As organisations race to operationalise AI, strengthen governance, and prepare for the regulatory shifts reshaping Europe, the leaders who will win in 2026 are those who turn today’s insight into tomorrow’s execution. Fast. This closing session distils the summit’s most important lessons across trusted data foundations, Responsible AI, regulatory resilience, data product operating models, and enterprise-grade AI scale.

Attendees will be guided through a structured reflection to identify one high-impact change they will implement in their data or AI strategy within the next 30 days - whether tightening lineage controls, accelerating a data-product rollout, strengthening model oversight, or redesigning cross-functional ownership.

By ending with shared commitments, this session transforms the summit from a day of ideas into a catalyst for measurable progress, equipping every leader in the room with a clear next step and the collective momentum to deliver it.

Aaron KalvaniGlobal AI StrategistUnited Nations