Executable, governed journeys
Customers don’t experience architectures; they experience journeys. But a great experience only happens if the journey can actually be executed — by people, AI agents and enterprise systems, governed end to end. Executable Journeys is the execution architecture of the customer experience: we design the journey, prove it in the Accountable Bank, and show the board the future working before it commits capital.
Most journey work stops at Customer → Screens → Success. A journey that has to run with AI agents acting for the customer needs the whole stack designed together — or it fails, quietly, in production.
The agent is almost the last thing you design. The market asks “where should we use AI agents?” The better questions come first:
Which business outcome are we trying to achieve? · Which decisions can be delegated? · Which need human judgement? · Which data products make those decisions possible? · Which controls must be embedded? · How do we generate the evidence a regulator will accept?
Whether a journey can survive those questions is an architecture problem — see the journey feasibility test. Reasoning is allocated in Cognitive and governed in Control.
We read every flagship journey at three levels: the future journey (what the customer experiences), the agentic execution (how it runs safely), and the architecture blueprint (how it is built and governed). These are board-level visions — Roadmap directions, not yet demonstrated.
Future journey: “We moved into our house in six weeks.” The winning bank may not be the one with the cheapest rate — it may be the one that gives customers certainty, speed and confidence at life’s largest decision. People don’t remember saving twelve basis points; they remember moving in.
Agentic execution — a cast of governed agents, with a human adviser in the loop:
Architecture blueprint: data products for customer, property, affordability, chain, identity and payment; controls and evidence on every agent action; chain-orchestration to completion certainty. How it’s built and governed →
Future journey: the bank becomes a trust layer — a reusable, consented digital ID that proves age, address and eligibility once, then opens accounts, mortgages and insurance without passports, utility bills and branch visits.
Agents: consent · attribute verification.
Architecture: identity & consent data products, trust framework, audit graph, credential wallet.
Future journey: the bank stops selling savings products and starts managing liquidity — reading spending, upcoming bills, risk appetite and tax position to recommend where money should sit. Where RPBI is heading.
Agent: financial-goals agent.
Options, all governed: savings, tokenised deposits, money-market funds, government bonds, ISA, pension — and DeFi only where suitable. Feeds Value.
The visions above rest on journeys already running on the Accountable Bank — on synthetic data, with test consumers, never a production service. Each is a tangible asset you can put in front of a board.
Relationship depth across a household, with delegated access kept under control.
Exercises: Trust · Cognitive · Control
Outcome: deeper relationships and deposit retention across the household.
Open a UK account before arriving in Britain — identity, funding and onboarding from abroad.
Exercises: Experience · Trust · Financial · Control
Outcome: acquire customers before competitors do.
Delegated authority over a high-value asset, without the customer losing control.
Exercises: Trust · Control · Cognitive
Outcome: serve complex, high-value clients safely.
An agent moves money between current and savings pots and covers a direct debit on the way back — so a customer never loses interest or trips an overdraft.
Exercises: Value · Cognitive · Control
Outcome: a proposition almost no one ships today.
An ordinary bank account for someone without a fixed address, documents or a smartphone — the bank accountable, the applicant in control.
Exercises: Experience · Trust · Control
Outcome: reach customers no one could serve before.
A payment experience designed to win the deposit, not just move the money.
Exercises: Value · Experience · Platform
Outcome: turn payments into primary relationships.
A transformation journey: legacy → shadow ledger → digital twin → parallel run → agent-assisted migration → evidence → switch → optimise. The existing core kept as one governed component.
Exercises: Platform · Control · Financial
Outcome: evolve the bank without a big-bang core swap.
An issuer moving to systemic scale under the Bank of England’s draft Code of Practice — backing, daily reconciliation, redemption and safeguarding.
Exercises: Value · Control · Trust
Outcome: scale new money under regulatory expectation.
Network-wide operations across many franchise businesses under one accountable structure.
Exercises: Platform · Control
Outcome: operate a network without losing accountability.
Where we’re taking the Accountable Bank next — and the conversations we’d start with you. Suggested directions, not yet demonstrated; bring one and we’ll prove it.
Every journey we demonstrate exists for one reason: to remove uncertainty before investment. The board sees the future working before it approves the programme — and leaves with a board-ready pack across four outcomes: customer (better for people?), operational (does it run?), regulatory (can we evidence it?) and commercial (does it create value?).
Tell us the outcome you want and who it has to work for. We will prove the journey, the agents, the authority and the controls before any of it hardens into build.
Tell us the proposition, architecture decision, or transformation problem you are facing. We will tell you, honestly, whether and how we can help.
Discuss a challengeLiving models and Field Notes on the future of banking. Published when the model changes, not on a schedule.