Organisations should stop treating AI as a box-ticking exercise to reassure their boards and instead start from ‘the reason, the moat and the product backbone’. That was the key takeaway from FintechOS Elevate 2025 this week.
Elevate 2025 was billed as FintechOS’s largest industry event to date, with more than 150 attendees from 50-plus organisations hitting the Pan Pacific London event venue for the end-to-end financial product management platform’s annual customer forum.
Kicking off the day, Teo Blidarus, CEO and co-founder of FintechOS, sketched out what he sees as the industry’s technology journey so far: from static legacy systems, to cloud-first systems, and now to what he called ‘AI-fluent systems’.
“The industry has moved from static legacy systems to cloud first systems,” he said. “Now it is switching to AI-fluent systems.” The catch, he added is that “we can’t achieve that shift with today’s product engines.”
That set up the event’s main announcement: FintechOS 8, described by the company as the world’s first Unified AI ProductOps Platform, aimed at helping banks and insurers manage the full financial product lifecycle with more speed, intelligence and control.
FintechOS 8
The platform – set to launch in Q1 2026 – marks a move to a multi-product lineup, with three main components:
- a ‘Unified Product Operations Center’, pitched as a command centre for each business line, aligning product strategy and execution and giving humans and AI copilots a shared view of products and performance
- an ‘Agentic Workforce’, a set of governed AI agents to take on work such as underwriting and claims management under transparent, auditable rules
- a ‘Data Core’, designed to decouple data from legacy systems and turn multiple data sources and integrations into reusable, governed data products
Each of these products is positioned as usable on its own, but the narrative across the day was about their value as a combined ProductOps backbone for banks and insurers looking to embed AI more deeply into everyday product work.
“To truly unlock growth with AI, institutions need a strong product operations backbone,” said Blidarus. “We’re here to help financial institutions modernise their legacy systems and innovate faster on a foundation of smart infrastructure and trustworthy AI.”
Panels on AI adoption
While the product launch drew attention, much of Elevate 2025 was devoted to how AI adoption is actually playing out inside institutions.
During the ‘AI Unplugged’ session, panellists from LHV Bank, Woodhurst, BRD – Groupe Société Générale and Advisority, joined Leda Glyptis to talk about the gap between board-level ambition and day-to-day delivery.
AI, they argued, is often set as the top strategic priority alongside existing change portfolios, resilience work and regulatory programmes, without the same level of clarity on scope or value.
A recurring theme was the idea of the ‘moat’. Borrowed from private equity, the question posed to attendees was simple: if AI investment is meant to deepen a bank’s moat, how exactly will it show up – in margins, cost, speed, risk or customer stickiness – and how will anyone know it is working?
We also heard warnings against AI theatre: pilots that exist mainly for slides, vendors without skin in the game, and ‘AI-powered’ labels that are impossible to verify. The advice was to anchor AI work in clearly defined product and customer outcomes, not in fear of missing out.
Leda Glyptis: 20 years of transformation, graded
In her keynote, Glyptis brought a more pragmatic tone of where financial services finds itself after two decades of ‘transformation’. As she put it, decision-makers have “been at the receiving end of an avalanche of stuff for the best part of 20 years” and, as an industry, “we have done okay… I’d give us a B minus. We have not failed but we have not aced this.”
That B minus crops up more than once in her assessment. “We need transformation and transformation is something we’re not good at. We get to B minus,” she said, pointing to research that suggests around 70 per cent of major programmes fail for predictable, often mundane reasons.
For Glyptis, the risk is that AI becomes “something we shall do” to answer the board’s question about what the organisation is doing on AI – but still without a solid foundation. “More than a vision you need a reason,” she argued. “And 90 per cent of those AI implementations that we’re seeing out there don’t have the reason other than this is a bit scary and they need to be a part of it.”
Her advice was to stop trying to learn everything from scratch and instead collapse the learning curve by trusting experienced people inside the organisation and working with partners who already have the expertise. “These are the surgeons already. They don’t need to go to school. And for everything else, bring partners and bring them in today,” she said – with a clear condition that those partners are held accountable rather than allowed to create new dependencies.
Glyptis closed on a simple reassurance: “You have everything you need in your organisation to navigate. It’s a question of finding the right people, getting out of their way and not repeating the mistakes of the past.”
Also on the agenda
Insurance featured prominently throughout the day, with speakers talking through how AI is being applied to areas such as claims, fraud and underwriting.
Ian Bubb, an independent consultant at Strategem Group, reflected on how quickly attitudes have shifted. “Even a few years ago, insurers were like, ‘this AI thing is not going to come’,” he said. “But now, with the organisation of cloud and restructuring of data, that’s actually how AI is becoming more real.”
FintechOS board member Rich Longo touched on everyday buying behaviour as the benchmark for financial products. “When you buy curtains on Amazon, at the bottom it says: you bought curtains, do you need a pole, do you need hooks? That’s a human buying behaviour,” he noted. “A majority of your buyers today for financial services buy that way – that’s the expected experience.”
Most of the interesting use cases were shown to rest on years of groundwork in cloud migration and data restructuring. Once data is available in the right shape, AI can help classify documents and images, surface suspicious patterns and refine pricing – but none of that happens without the less glamorous integration and governance work.
This is where FintechOS’s Data Core and Agentic Workforce pieces were positioned, with demos showing AI agents working on top of datasets rather than delivering any kind of magic-wand automation.
