Travel Ecosystem

A 9-week discovery sprint and 2030 vision for the bank as the financial layer of travel — AI-agent orchestration included — that initiated new delivery projects on the 2026 roadmap.

The Problem

Travel is one of the largest and most fragmented spending categories for the bank’s customers, and almost none of that spend happens inside a banking relationship — it happens across airlines, agents, and a growing layer of AI travel agents booking and paying on customers’ behalf. Leadership wanted a point of view: what role, if any, should the bank play in travel by 2030, and what should it start building now.

The brief: define where the bank can be genuinely useful in travel by 2030, without pretending to become a travel company.

My role / the team

I led a 9-week discovery sprint as design lead, working with a strategy partner and two engineers prototyping AI-agent scenarios. The sprint output — a 2030 vision and a set of near-term bets — directly initiated new delivery projects now on the 2026 roadmap.

What The Evidence Said

Expert interviews across travel and fintech found that AI agents will handle a growing share of travel booking and payment decisions well before 2030. Customer research showed people trust their bank with money more than with itinerary judgment. And a landscape review found competitors either ignored travel entirely or over-invested in becoming travel companies themselves, with little in between.

The Judgment Calls

‘Brilliant at the financial journey’ — letting others be brilliant at travel.

The instinctive pitch was to build a travel app. I scoped the vision away from that, to owning the financial layer only — payments, protection, currency, and agent-verified transactions — because trying to out-build dedicated travel platforms would have spread the bank thin at exactly the moment AI agents are commoditising the booking layer anyway.

An augmentation layer for AI agents, not a bank-owned travel agent.

Rather than building a competing AI concierge, I framed the bank’s role as the trusted financial layer underneath third-party and AI travel agents — verifying, authorising and protecting transactions made on a customer’s behalf. It’s a smaller, less visible role, but a far more defensible one.

Designing for 2030 agents without losing 2026 feasibility.

It would have been easy to produce a vision only buildable once AI agents matured. I anchored every 2030 bet to a smaller, shippable 2026 version — so the roadmap could start now rather than wait for a future that hadn’t arrived yet.

The Design

The financial-layer model — where the bank sits relative to agents and travel platforms.

An agent-authorisation concept, one of the 2026-feasible near-term bets.

The 2030 vision map, sequenced back to a 2026 starting point.

What Shipped & What Happened

The discovery sprint concluded with leadership sign-off on the financial-layer positioning, and it directly initiated new delivery projects now entering the 2026 roadmap. There’s no shipped product yet — the honest measure of success at this stage is that the bank now has a funded starting point instead of an open question.

What I’d Do Differently

I’d bring a payments-platform engineer into the sprint itself rather than after — some of the agent-authorisation bets look simpler in a vision deck than they’ll be to actually build, and earlier technical grounding would have sharpened the near-term bets.