Find where AI can improve the way your team works.
We map the workflow, identify a worthwhile change and help your people put it into practice. You get a clear starting point, an agreed plan and a way to measure the result.
A 30-minute conversation to understand the work and whether an audit is the right next step.
Have a licence-spend question? Estimate the waste in 2 minutes
Input
An invoice arrives by email, attached to a message.
What AI does
It reads the invoice, matches it to the right contract and drafts the reply.
Human check
The person who owns the contract checks the match and approves it.
Outcome
The reply goes out and the record is filed against the contract.
An example of the shape, not a client result. Your audit maps your own work, step by step.
Recognisable work
Where this usually helps first.
Three situations we are asked about most often. They are written as hypothetical examples to show the shape of the problem — not as descriptions of any client.
Sales
The handoff loses the context.
What was promised on the call sits in someone’s notes rather than in the system, so the next person rebuilds it from memory — or asks the customer again.
Operations
The same figures, typed twice.
Numbers are entered into one system, then into another, then reconciled by hand — because the two were never connected and nobody owns the join.
Administration
An accurate summary that still leaves work behind.
The meeting notes are correct. Someone still has to check the decisions, name the owners and move the tasks into the system people actually use.
Hypothetical situations, written to be recognisable. Not client claims.
The tools didn’t fail. And neither did your people.
Nobody had ever seen AI do their own job. Show them once, on the work they actually have in front of them, and this is what comes back: “This is amazing. I didn’t know that it can do this. The work I was doing for three hours right now takes 10 minutes.” — a specialist at an energy systems group, after hands-on sessions with their team.
How an engagement works
What we look at, what you receive, and what your team does.
One workflow at a time, starting with the people who run it. We are vendor-neutral: we sell no licences, so a recommendation to cancel a tool or leave a process alone costs us nothing.
What we investigate
- How the work moves: each step, each handoff, and where information comes from.
- Which steps are still done by hand, and how long they actually take.
- Which tools and AI seats you pay for, and who opens them.
- Which decisions should stay with a person.
What you receive
- A written map of the workflow as it runs today.
- A measured starting point, in hours per week.
- A prioritised plan: what to change first, what to cancel, what to leave alone.
- An EU AI Act Article 4 chapter — the AI-literacy foundation, not compliance advice.
How your team participates
- Short interviews with the people who run the process — usually 30–45 minutes each.
- One handover of licence and usage data from whoever administers your tools.
- A sponsor who can decide what changes, and a named owner for the workflow.
- Hands-on sessions where the team practises the new way of working.
The steps, in order
- AI Workflow AuditPaidfrom €2,000
The paid first step. We map one team’s workflow end to end, measure the manual steps and hand back the Workflow & Usage Report with a prioritised plan.
Learn more - Enablement WorkshopsPaid
One real workflow changed hands-on with the team that runs it, inside the tools they already have.
Learn more - AI Enablement ProgramPaid
Your champions run the day to day; we stay as the architect and report the hours back each quarter.
Learn more - AutomationPaid
Once the team already works the new way — we build the automation, or teach your champions to build it.
Learn more
Always the audit first — it sets the starting point and the plan. Where the audit recovers licence spend, that money can fund the work that follows.
Free — a licence-spend estimate, in two minutes
Five questions about what you pay for and how many people use it, and an estimate of the licence waste on screen. It estimates licence spend, not workflow hours, and it is not the audit.
Open the free Self-CheckThe people doing it
Who you actually work with.
Senior-led: the two of us map the work and sit with your team. No handoff to juniors.
Software engineering, team leadership and teaching, plus running a business herself — CEO, CTO and operations work. She designs the change with the people who will run it, and fixes the process before automating it.
Connect on LinkedIn
Nine years in enterprise systems and SAP S/4HANA, at Generali, Carrier and Zebra. He reads a system the way the people who build it do, and knows what an integration will actually cost you.
Connect on LinkedInApproved proof
One client engagement, measured — and what it does not show.
This is the work we have written permission to publish. The client is anonymous permanently, and every figure below is theirs, confirmed by them.
Client story · an energy systems group
to write one meeting protocol, by hand
after the change
for the same protocol — 289 of them in six months
Four processes now run with a check instead of by hand — about one full-time role of manual work, out of the month.
An energy systems group asked us to look at what its people were still doing by hand. We talked to them, mapped the work, then built it with them: tender-document assembly, invoicing, email and correspondence handling, and meeting capture — alongside hands-on sessions on working with Gemini. Over six months, 289 meeting protocols were written at about four minutes each instead of about an hour: roughly 44 hours a month back.
The correspondence work held the same standard: 96 incoming letters in six weeks, all 96 matched to the right contract. Taken together, the four processes remove about one full-time role of manual work a month.
About four minutes instead of about an hour per meeting protocol, across 289 of them in six months.
96 incoming letters in six weeks, each matched to the right contract — 96 of 96.
Across the four processes, about one full-time role of manual work a month — hours freed, not a job removed.
- One client, four processes, over six months. It is not a benchmark for your business.
- The figures are in hours. There is no € figure: converting hours to money needs a cost per hour we do not assume.
- Nothing here runs unsupervised. Each process still has a person who reviews the output and owns the result.
Before you decide
The questions leaders ask us first.
How much of our team’s time does this take?
A few hours in total for a Scoped audit. Short interviews with the people who run the workflow — usually 30–45 minutes each — one handover of licence and usage data, and a walkthrough at the end. The workshops are scheduled around your week.
What access do you need?
For the audit: conversations with your team, and a read-only export of licence and usage data from whoever administers your tools. Where a later change needs an AI system to read or write in a real system, we agree in writing what it may do before anything is connected. A standard NDA comes first — yours if you have one, ours if you don’t.
Will this work with the systems we already have?
That is part of what we investigate. We are vendor-neutral and build on the tools you already own; where a system cannot support the change, we say so and propose the simpler option instead of a new purchase.
What is in scope, and what is not?
A Scoped audit takes one team or department in about a week. A Full audit covers the whole company over 2–3 weeks. We do not give legal or compliance advice: we build the AI-literacy foundation the EU AI Act Article 4 duty asks for, and refer the legal reading elsewhere.
Who owns what you build?
You do — the report, the plan, the automations, the trained champions, all in your accounts and your people. If you stop working with us, it keeps running. If you continue to the enablement program, the audit fee credits toward it.
What does it cost, and who qualifies for the 3× guarantee?
A Scoped AI Workflow Audit is €2,000–3,000 for one team; a Full audit is €6,000–12,000. The 3× guarantee applies above a qualification floor — roughly €40,000 a year of visible SaaS and AI licence spend, or about €6,000 of estimated annual waste. Below that we route you to the Small Audit instead, because the arithmetic behind the guarantee cannot reach 3× of the fee. Full terms are on the audit page.
Start with one workflow.
Bring the one your team dreads. Thirty minutes, no slides: you describe the work, we tell you what we would change first and whether an audit is worth doing at all.
A 30-minute conversation to understand the work and whether an audit is the right next step.
Licence spend rather than a workflow? Estimate the waste in 2 minutes
- You own everything we build
- Vendor-neutral — no lock-in
- Senior-led — no juniors
- EU AI Act Article 4 foundation
No follow-up unless you ask for one.