How to Implement AI Successfully: Lessons from London Tech Week 2025

This year's London Tech Week delivered something we don't often see in the tech conference circuit—honest conversations about what actually works when it comes to AI implementation. Our team came back energised, not by flashy demos or vendor pitches, but by real stories from leaders who've been in the trenches of transformation.

Start with the Need, Not the Tool

Perhaps the most powerful moment came from Lakshman Natan, CIO at Paramount, whose keynote tackled the elephant in the room: AI fear factor. His message was refreshingly straightforward: Start with the need, not the tool. What are your real use cases?

This isn't just good advice; it's essential. We've seen too many organisations rush into AI initiatives because they feel they should, rather than because they've identified a genuine business need. Natan's emphasis on bringing IT and business together under leadership alignment struck a chord with us. As he put it, "Each business unit has different learning patterns and paths. Tech and IT can therefore not operate in silo."

Align with your business on the AI journey, otherwise it's just an IT project.

Tech for Good, Not Tech for Tech's Sake

Shobie Ramakrishna from GSK reinforced this theme with her own compelling perspective: "Don't just do tech for tech. I'm only interested in tech that really changes and transforms." Her approach of embedding individual tech objectives into everyone's agenda—including the C-suite—creates both accountability and creativity in adoption.

This resonated deeply with our own experience. The most successful transformations we've supported aren't driven by technology departments in isolation, but by organisations that embed transformation objectives across all levels of leadership.

The Human Element Remains Central

One of the most interesting insights came from discussions about AI in financial services. Despite all the advances in AI capability, speakers consistently returned to the importance of human interaction. As one panellist noted, "Responsiveness, empathy, clarity: it always comes down to a human interaction with a machine."

This is particularly relevant for organisations implementing customer-facing AI. Starling Bank's launch of a generative AI tool that allows customers to query their banking activity using natural language ("How much did I spend on coffee last month?") represents a thoughtful approach—putting AI directly into customers' hands in a way that feels natural and helpful.

Real Users Will Surprise You

The BBC's Storm Fagan shared something that every organisation embarking on digital innovation should remember: "Real users (vs synthetic) will surprise you." There's genuine value in spending time with actual users rather than relying solely on theoretical models or synthetic data.

This insight connects to a broader theme we observed throughout the week—the most successful AI implementations start with deep understanding of customer needs and work backwards from there. David Schwimmer from LSEG reinforced this approach, emphasising that it's essential to work backwards from customer needs to guide development and innovation.

Building for the Future While Solving Today's Problems

What struck us most about the conversations was how forward-thinking leaders are balancing innovation with practical delivery. Lisa Jacobs from Funding Circle spoke about building "a modular, tech-agnostic stack to remain flexible as AI evolves" while maintaining focus on delivering great customer experience today.

This balance between future-proofing and present-day value creation is something we see in the most effective transformation programmes. It's not about implementing the latest technology because it exists; it's about creating foundations that enable continuous adaptation and improvement.

The Leadership Imperative

Perhaps the most critical insight from the week was about leadership's role in enabling transformation. David Buttress from OVO emphasised that "CEOs today must embrace greater risk and move quickly," while also stressing the need to "separate core operations from innovation teams to avoid slow-down from process-heavy environments."

This separation—protecting innovation from operational drag while ensuring alignment with business objectives—is crucial for organisations serious about transformation.

What This Means for Your Organisation

The conversations at London Tech Week reinforced something we've long believed: successful AI and digital transformation isn't about technology, it's about people, process, and purpose working together effectively.

If you're considering AI implementation in your organisation, start with these questions: