We sit on your side of the table, then hand you the answer as working software, financial models and operating change your team can actually run, not a deck. A proven ecosystem of platforms, tailored module by module, and run for you after launch.
How we work
Most consultancies hand you a deck and leave. Most software firms hand you a licence and leave. We do neither: we work out what your business actually needs, build exactly that, and stay to run it.
You do not need a steering committee or a data strategy. You need the jobs that eat your week to stop eating it, and one system instead of six subscriptions that refuse to talk to each other.
Where AI genuinely moves your numbers, what to buy rather than build, and whether you need it at all. If the honest answer is that your problem is a process rather than software, we say so.
Start a conversation → 02 · TailorWebsite, CRM, bookings, documents and follow-up wired into one place, composed from parts that already work for other businesses. Where a piece does not exist, we build it for you.
See the ecosystem → 03 · DesignNot a training course you hope sticks. We work out how the job should actually run, then build the system that makes it run that way. Tracking safety certificates and visas across a large workforce becomes a dashboard that surfaces what is expiring and who to chase, rather than a spreadsheet somebody has to remember to open.
See a real build → 04 · RunIt ships on your own domain wearing your brand, and we host it, update it and add to it as you grow. Nothing to patch on a Tuesday night, and no plugin stack to babysit.
See it live →Advisory that has to survive contact with audit, regulators and the operating reality of the business. We work alongside your operating, technology and risk leadership, inside the constraints you already have.
We sit on your side of the table first: where AI genuinely moves your numbers, what to build versus buy, and a roadmap your team can execute. The deliverable is one your finance and risk functions can sign.
AI advisory → 02 · ProveMost of the argument is about gaps: what the current stack does not cover, and whether to fill it with a product, an integration or something built. We work that through with rapid prototypes, so the decision rests on a working thing your team has used rather than a vendor demo or a scoring matrix.
How we assess → 03 · EmbedThe technology is rarely what fails. We handle the change management around it: process redesign, decision rights, the governance a regulator will ask about, and the ways-of-working that get people actually using it. Treated as a delivery outcome with owners and measures, not a training day at go-live.
Start a conversation →AI advisory
The hard part of enterprise AI is not the model, it is everything around it. Most failures inside large organisations are governance, integration and change failures. A model that scores 0.94 on an internal benchmark and never reaches a customer is not a success, it is a sunk cost waiting to be written down.
Translate AI capability into an investment case your CFO can defend, in the language your finance function already uses.
Which processes carry enough cost or risk to be worth automating, and which are simply irritating. The second list is usually longer.
Costed honestly, including the run cost nobody models: retraining, monitoring, licences and the people who own it.
Written so your finance and risk functions can interrogate it, because they will, and so the board can sign it.
Production systems, not pilots that quietly die in Q4 once the enthusiasm and the budget run out.
Into the CRM, the document estate and the finance systems that hold the real data, rather than a demo standing to one side of them.
Authentication, permissions, logging and monitoring, solved on day one. This is where most pilots quietly fail.
Your own people can operate it, extend it and explain it after we step back. If they cannot, it was never in production.
Model risk frameworks designed alongside legal, security and compliance, from the first architecture decision rather than at the review before launch.
Audit trails from an output back to the prompt and the source document, so a regulator or an auditor can follow the reasoning.
Where your data goes, which models see it, what is retained and what crosses a border, written down before anything ships.
Human oversight, continuous risk management and event logging aligned to the EU AI Act, built in rather than retrofitted.
AI systems decay quietly. Prompts drift, data shifts and usage tails off months before anyone reports a problem.
Telemetry on what is actually being used and by whom, so the difference between adopted and abandoned is visible early.
Prompt and model maintenance sits with a named owner on a defined cadence, not with whoever happens to notice.
Each engagement leaves the capability stronger than it found it, so the investment compounds instead of flattening.
Twenty-five years directing large, complex change where the stakes are structural — separating regulated banks, carving finance systems out of global mergers, and standing up operating models from nothing.
The hardest part of any deal is the moment one organisation becomes two, or two become one, without interrupting trade. Cutover planning, interface and dependency mapping, legal entity structures and shared-service standup.
Owned the integrated cutover splitting investment banking from retail banking and stood up a new shared services company alongside them. Three ERP modules, 100+ interfaces, £5bn in assets, 70,000 employees and roughly 100,000 cost centres — where a dropped data link means someone does not get paid. Both banks kept trading.
~£1bn programme · UK regulated banking
Programme director through one of the largest transatlantic mergers in corporate history: a global front-to-back office and accounting solution across 600+ legal entities, then the clean separation of finance systems and structures at demerger, including restructuring regional treasury centres.
£200m programme · 100+ FTE delivery
Designed the separation programme framework and delivery governance across organisational, technology, finance and regulatory dimensions, establishing reporting cadences and decision forums that held pace without excess bureaucracy.
Regulated captive finance
Designing how an organisation actually runs — reporting lines, accountabilities, decision rights, end-to-end processes and handovers — then implementing it. Equally at home merging functions together or building a structure from a standing start.
Repositioned treasury from cost centre to value-creating function across a private equity real estate portfolio, and led design of an in-house banking model with cost allocations, recharges and financing across waterfall structures.
$300bn AUM · 10,000+ bank accounts
Merged risk, liquidity, funding and covenant management into a single integrated function and platform for a European private equity house, defining cross-functional processes, handovers and decision rights throughout.
PE-backed transformation
Built a full multi-jurisdictional target operating model, governance framework and technology platform from a standing start: reporting lines, decision rights, steering committee governance and an investor-grade reporting cadence.
Multi-jurisdictional scale-up
Standing up governance where none exists, and providing independent assurance where delivery is already underway. Built to create pace, not paperwork.
Decision rights, RACI, RAID, steering committee cadence and benefits tracking established from nothing, then driven into daily use through ways-of-working playbooks rather than left as a document set.
Portfolio & programme level
Built the global programme management framework for a financial software and consulting firm, adopted as standard across its entire consulting division.
Consulting division standard
Assurance over vendor-managed implementations across private equity and global holding groups: validating model outputs, reconciliations and reporting, and running steering forums across finance, treasury, tax and IT. Reports what is actually true about a programme, including when that is unwelcome.
Multi-vendor environments
Programmes that have stalled, lost a vendor, or drifted out of regulatory tolerance — diagnosed for what is genuinely broken, then rebuilt around a delivery path that holds.
Recovered a collapsed regulatory reporting workstream by independently designing and building a bespoke EMIR reporting solution, restoring compliance without re-tendering.
EMIR · regulated finance
Designed and implemented a SOX-compliant application management and security framework, subsequently adopted and reused by a global industrial group, a central bank and a European energy utility.
SOX · reused across three institutions
Delivered the Solvency II data extraction capability from an actuarial modelling platform for a UK insurer, meeting the regulatory compliance deadline.
Solvency II · insurance
Putting AI into an operating model with the controls a board and a regulator can accept — approval zones, graduated autonomy, human oversight and audit trails — rather than as an unsupervised experiment.
Designed and operated a multi-agent governance framework: six specialised agents across data, marketing, technology, commercial, finance and QA, orchestrated on weekly OKR cycles under a green / amber / red approval structure with autonomy levels 0 to 4 gating every external action.
Live operating framework
Designed to align with the EU AI Act's core obligations: human oversight, continuous risk management, automatic event logging and transparency to deployers — designed in from the start rather than retrofitted.
Governance by design
Designed and built seven production systems directing AI coding tools, including a GRC and knowledge platform with immutable audit trails, an M&A and investor data room, a self-service financial modelling portal and a private-market tokenization platform with on-chain KYC controls.
Architect and developer
The part most programmes underfund: getting people to actually work the new way. Treated as a delivery outcome with owners and measures, not an afterthought at go-live.
Designed structured organisational change frameworks — change maturity, innovation maturity and reward and recognition — taken up and run by financial services clients as their own.
Organisational change design
Built the operating model, automated onboarding, structured change communications and ways-of-working playbooks that let a workforce scale by roughly 500 staff per month without the operating model breaking.
~500 joiners / month
Deployed a QA reviewer agent that checks every client-facing artefact against source-of-truth files for factual accuracy, tone, approval-zone compliance and internal consistency before it is sent.
Automated pre-send review
Client organisations are described by sector and scale rather than named. Details available under NDA.
Our business model
Three tiers that let us put senior people on the problem without carrying the overhead of a large firm, and scale up or down as the work demands.
Battle-hardened senior executives with deep knowledge of their vertical markets, who have managed global change and delivery models across start-ups, scale-ups and multinational organisations and institutions.
Talented, dedicated consultants, subject matter experts, technical associates and highly experienced industry-leading practitioners.
Innovative, trusted partners who help us create solutions that lead the market in flexibility, interoperability and client value, enabling us to stay lean yet exceptionally scalable as client demand grows.
Applied Intelligence, in writing
Field notes from running AI inside real businesses. No hype, no doom, just what works and what quietly fails.
What actually changes when an AI workforce reports into a human chief executive.
Read it → NewsletterWhy AI initiatives die walking, and the early signals that yours is already undead.
Read it → NewsletterAI does not crash. It returns a confident, well-formatted answer that happens to be wrong.
Read it → ArchiveThe full archive, including print publications and features.
Browse the archive →Start a conversation
Tell us how your business runs and where it hurts. We will tell you, honestly, whether AI moves it, what we would build, and what that takes. If the answer is no, you will hear that too.
Booking runs on our own CRM's Online Booking module, the same one we ship. Prefer email? We reply within one working day.