Find where AI actually pays off in your operation.
Leave with the builds.
Reframe analyses how your business really runs and pinpoints the steps where AI and automation return real money, verified by your own people and priced per step. You leave with a build-ready brief, or a working MVP, for every opportunity you select. That takes weeks. A consulting scan takes months.
Every route to AI runs between two bad ends.
You have seen both pitches. The army of consultants, and the tool that claims AI solves everything. Both fail the same test: show where the money is, prove it, then build the fix.
The consulting army
Twelve people, six weeks, hourly rates and a 200-page deck. Serious machinery, and most of it is pointed at producing presentations. The overhead is the product, and the insight leaves with the team.
The AI wrapper
A slick demo on top of a general-purpose model. Fast, cheap and confident, and blind to how your operation actually runs. AI is a powerful instrument. It is a poor substitute for experience, context, and judgment.
[re]frame was built to sit deliberately between the two: senior operators with decades in process and change, running an engine we developed for this one job. Small team, machine speed, verified output, and a build at the end. See the method →
From how you work today to the AI that should improve it, and the build to start on Monday.
Five steps, one loop. The system analyses at machine speed. Nothing advances until the person who owns the process signs it off, and nothing gets recommended without the ROI arithmetic to back it.
Map the operation
How the work actually runs, mapped with the people who do it: every hand-off, decision point and dependency on people, departments and tools.
Sign-off loop
Every AI-generated map goes to the department head who owns it. Signed off, it proceeds. Rejected, it goes back. Only validated maps move on.
What each step costs
The current cost of every step, built from role-specific rates, volumes and rework, and the saving on the table if technology takes the work over. Every assumption stays editable.
Find where AI pays
Each step scored for AI potential against what it costs today, ranked by return, with the arithmetic shown.
Briefs & MVPs
Each selected opportunity leaves with a build-ready brief or a working MVP, plus the quantified roadmap and board deck.
Why you can trust what comes out.
The honest objection to AI-led analysis is garbage in, garbage out. That objection is why the loop exists: a human signature sits between every input and every recommendation, so what reaches your boardroom stands on a truth your own department heads have signed.
Process notation
Standard notation is its native language. Maps your architects recognise.
Dependency model
People, departments and tools, mapped per step.
Cost engine
Role-specific rates and editable assumptions, per step.
Verification workflow
Every output routed to the person who owns the truth.
Six artefacts.
From verified reality to buildable AI.
Verified as-is map
How the operation actually runs, in standard notation your architects and auditors will recognise, with every dependency on people, departments and tools made visible.
Verification log
A record of who reviewed and signed off each map and analysis. This audit trail is what makes every downstream AI recommendation defensible.
Per-step cost model
Role-adjusted cost per step with editable assumptions, rolling up to the true cost of the operation as it runs today. This becomes the baseline every opportunity is measured against.
AI opportunity map
Every step scored for AI and automation potential, colour-coded by optimisation archetype and ranked by return, so selection becomes a reading of the numbers.
Quantified roadmap
Selected opportunities sequenced by payback and dependency risk, with the savings arithmetic shown in FTE hours, cost reduction and euro impact, plus the board-ready deck in one click.
Build briefs & MVPs
A dev-ready brief covering scope, data, integration points and KPIs for every opportunity you select, or a working MVP where you want proof before you commit.
The system is half the method.
We're the other half.
AI tooling like this is everywhere right now, and in the wrong hands it's vaporware. Rapid scanning without context, without knowing how a process owner thinks or what change takes inside a large organisation, produces fast, confident nonsense. Reframe is run by three senior business leaders who have spent their careers in operations, change management and consulting. We frame the diagnosis, the system scans and quantifies at machine speed, your process owners verify the findings, and we judge what will survive implementation.
Yassine el Kochta
Over 15 years advising CIOs and CDOs on operating models and AI adoption. Led an innovation and AI unit for 11 years, delivering 60+ programmes, and ran process and governance work across a major European banking group.
Bas Redhead
Senior commercial executive with 25+ years in large international organisations. Builds AI products hands-on, and has spent over a decade as an independent strategist helping organisations sharpen their story, their marketing and the process behind both.
Rutger van der Lee
AI, digital and omnichannel specialist in pharma. Founded an AI academy that trains commercial teams, has facilitated business wargames for 18 years, and ran his own strategy and business development practice for 15.
HOW IT RUNS IN THE ROOM · Seniors frame → the system scans & quantifies → your owners verify → seniors judge. Human expertise and machine speed, hand in hand. Neither is credible here without the other.
Three routes to an AI roadmap.
One survives scrutiny.
One end of the market sells hours. The other sells a wrapper around someone else's model. Both leave the same gaps: no verified ground truth, no defensible numbers, and nothing you can build from.
| The consulting army | The AI wrapper | [re]frame | |
|---|---|---|---|
| Built on | Hours, headcount and interview notes | A general-purpose model with a chat window | An engine developed specifically for operational diagnosis |
| Grounded in | Workshops and maturity models | Whatever you type into it | A verified map of how work actually runs, signed off by your own department heads |
| Opportunities chosen by | Workshop consensus and partner intuition | Whatever the demo makes look easy | Per-step economics with the ROI arithmetic shown |
| Cost | Six-figure fees, most of it overhead | Cheap, until you act on its advice | A fraction of a consulting engagement |
| Delivered by | A leveraged pyramid of juniors | A tool, on its own | Senior operators who have run businesses, processes and change, working hand in hand with the system |
| Ends with | A deck and an invoice | A demo that never survives contact with your operation | Build-ready briefs and working MVPs |
Start with one process.
End with your first AI build.
Reframe is delivered with our team in the room first. Once the diagnosis has earned your trust, you decide how far the building goes.
Pilot diagnosis
One core process, end to end: verified as-is map, per-step economics and a first shortlist of AI opportunities with the ROI arithmetic. Enough to judge the method on your own numbers.
Entry engagement
The full diagnosis across a department or value chain: verified as-is, the AI opportunity map, a quantified roadmap, build briefs for the steps you select and the board deck, with KPIs defined up front so the outcome is measurable.
Build, scale or license
We build the MVPs, roll the method across further processes, move you onto licensed access so your own people run it, or index fees to realised savings once the numbers are proven.
Your brand, our engine.
Reframe can be licensed to consulting teams who want to deliver verified AI-opportunity diagnosis, and the builds that follow, inside their own engagements at a speed and margin their current scan work can't reach. Partnerships are limited and selective while we scale.
Pick one process.
We'll show you where AI pays.
A 30-minute call: you bring the process you suspect is bleeding money, and we walk you through what a pilot diagnosis and the first build after it would look like.