Four signals, one answer
Over the past few weeks, four signals came together that point in the same direction. The AI Act starts enforcing in August. About half of employees use AI outside of policy. AI agents are going mainstream in enterprise software. And a large majority of SMBs have deployed AI, while nearly half do not know how to use it effectively. Four different stories, one conclusion. Before you build further, you need to know what works for you and what does not.
This blog describes exactly what an AI audit is. Not as an idea, as a product. What is in it, what you get, what you do with it.
What the audit is not
Managing expectations helps before we begin. The audit is not a technical compliance check. Not a vendor comparison. Not an 80-page PowerPoint that disappears into a drawer. And not a sales pitch dressed up as a report.
The audit is also not meant for micro freelancers, pre-product-market-fit startups, e-commerce with thin margins or government organisations with ten decision layers. It is built for growing professional-services firms of 15 to 75 FTE where the back-office team is stalling on recurring reports, files or customer communication.
What the audit is
Fourteen calendar days from the interview day to the report. Two weeks of lead time. Eight to ten hours of time investment on your side, spread across interviews and validation. A fixed price up front, no after-billing.
In those two weeks, we systematically map where AI can win time and quality in your company. Not based on gut feeling, but on interviews with you and four to six employees, observation of the work that comes in daily, and a classification by impact and feasibility.
The five concrete deliverables
One, an AI opportunity report. Between 8 and 15 opportunities, scored on impact and feasibility. Per opportunity, a short description, the expected time benefit, the data required and the dependencies.
Two, a process inventory. An as-is overview that sets out where the time, frustration and repetition currently sit. Per department. Per type of work.
Three, a roadmap. The top 3 to 5 opportunities worked out, with order and dependencies. Which process first, what is the build time, what effect to expect, which integrations are needed.
Four, a closing presentation. A session with the owner-director and the team where we go through the report step by step, answer questions and confirm the priorities. The slide deck stays in your possession.
Five, a follow-up proposal. A concrete quote for implementation or a retainer, based on the top priority from the roadmap. No obligation to continue, but a clear picture of what the first step costs and delivers.
How it runs over two weeks
Days 1 to 3, kickoff and interviews. We speak with you plus four to six employees from the departments where the work stalls. Back office, customer communication, planning, possibly finance. Each conversation lasts 45 to 60 minutes. We probe on time, frustration and repetition.
Days 4 to 10, analysis and drafting the report. We process the conversations, look at your systems where needed, follow up on specific processes and draft the opportunity report and the roadmap.
Days 11 to 14, finalising and the follow-up proposal. Report finished, presentation ready, follow-up proposal worked out based on the priority that emerges from the analysis. On day 14 or shortly after, the presentation to you and your team is in the calendar.
The official promise
In 14 days, you get a crystal-clear picture of where AI can save at least 50 to 100 hours per month in your company, plus a concrete roadmap to realise this in 6 to 12 months.
No unlimited. No 10x. But an honest, repeatable figure that we see recurring in earlier audits at service providers of your scale.
Two examples from practice
Back-office example. A mortgage firm with 35 employees processing 60 financing requests per day. The audit surfaced four opportunities, of which classifying and pre-filling the CRM based on incoming files had the highest impact. Expected time saved, around 70 hours per month across the whole team. The follow-up proposal was a six-week implementation.
Customer-communication example. An estate agency with four branches where the approval process for social posts ran slowly. The audit found four different WhatsApp groups, two approvers and two Excel files for planning. The follow-up proposal was a central approval dashboard with brand-style checks and publication planning, and with it around 25 hours per month less coordination.
Not the very biggest figures. But the kind of opportunities we see structurally.
What comes after, if you want it
The audit is no obligation to continue. Three paths run in parallel afterwards.
One, you get to work with the roadmap yourself. Your own pace, your own priorities. We remain available for questions.
Two, you have us implement the first use case in five to eight weeks. Fixed scope, fixed price based on the roadmap.
Three, you go straight to an AIOS retainer in which we stay structurally connected as a partner and keep developing the AI layer beneath your organisation.
Which path fits, we discuss during the closing session. No pressure to decide today, but clear information to walk away with.
Who this audit is not for
An honest line in the sand. Not for micro freelancers and sole traders. Not for pre-product-market-fit startups. Not for e-commerce with thin margins and no back-office pain. Not for government and non-profit, because decision-making and budget do not fit this format.
It is for professional-services firms with repetitive back-office work, a working business model, and an owner-director who does not want to wait until a competitor has it sorted first.
The first step
If you recognised one of the four signals in the past few weeks, the AI Act, shadow AI, failed pilots or teams working hard without getting ahead, then this is the moment.
Fourteen days later, you know where AI works for you, what it delivers and what the first step is. AI that works in your field starts with knowing what is possible at your company.
