01
Data your models can't reach
Records locked in legacy databases, PDFs, and vendor systems with no clean interface. Until that data is accessible and trustworthy, every AI pilot dies at the demo stage.
Division 06Digital Transformation Consulting
Close the gap between legacy systems and modern AI.
Most enterprises don't have an AI problem. They have a plumbing problem — decades of systems that can't hand data to each other, let alone to a model. This is the work of fixing that, in production.
Digital transformation consulting has a reputation problem: expensive diagnosis, thin delivery, and a roadmap nobody can execute. The engagements here are built the opposite way round. We start from the operational bottleneck that costs you money — a manual reconciliation, a 12-day onboarding cycle, a support queue nobody can staff — trace it back through the systems and data that produce it, and ship the smallest thing that removes it. Then the next one. Legacy ERP, ageing CRMs, spreadsheets holding the business together: none of that has to be ripped out first. It has to be made legible to modern tooling so AI, automation, and analytics can finally reach it.
What I build
Where enterprise transformation actually stalls.
02
Process nobody has mapped
Workflows that exist only in people's heads. Automating an undocumented process just makes the mess faster. We map first, then cut.
03
Pilots that never reach production
Proof-of-concepts built outside security, compliance, and operations. We design for the production path from day one — auth, audit, monitoring, ownership.
04
No internal capability to hold it
Systems handed over without documentation or trained owners decay in a quarter. Enablement is part of delivery, not an upsell.
How I work
How an engagement runs.
- 01
Diagnostic (2–4 weeks)
Interviews, systems inventory, and data-flow mapping across the functions that matter. Output: a ranked list of bottlenecks with cost, effort, and risk attached.
- 02
Roadmap tied to outcomes
A sequenced plan where each phase names the metric it moves — cycle time, cost-to-serve, error rate — and the systems it touches. No transformation theatre.
- 03
Ship the first capability (90 days)
Integration layer, automated workflow, or AI service built into your environment, through your security review, used by real staff.
- 04
Scale and hand over
Repeat in six-week increments, with documentation, runbooks, and training so your team owns what we built.
Selected work
Who this is for.
Operations leaders
Cost trapped in manual work
Reconciliation, data entry, and handoffs consuming headcount that should be doing judgement work.
Technology leaders
Legacy estate blocking AI
A modernisation mandate, a real budget, and systems that predate the tooling everyone expects you to adopt.
Founders & executives
Pilots that never landed
AI initiatives that impressed the board and changed nothing operationally. This engagement is judged on shipped systems.
Next step
Start with the diagnostic.
Bring the bottleneck that costs you the most. We'll map what's producing it and what it takes to remove it — before committing to anything larger.