I audit one area of your business, name the places where generative artificial intelligence would save you time, then build the systems with you. You take back the controls whenever you want.
How it runs
Diagnostic
10 days
Pilot build
4 to 6 weeks
Industrialisation and handover
fixed fee or monthly
You decide whether to continue after each stage.
×15
traffic at a 20-person consultancy, from 10 to 150 visitors a day
×6
inbound enquiries, from one every three days to two a day
1 in 4
of those inbound enquiries turns into signed work
Figures measured on an acquisition build running since March 2026 for a 20-person consultancy.
Read the case studyThe difference
It is that nobody in the building has time to work out what ought to be automated, or to build it. A training course answers neither question. A system that runs answers both.
The diagnostic starts from the time your teams actually spend, not from a catalogue of use cases. If an area holds no gain, I tell you so.
An AI course goes stale within months and still never says what to automate here. Build the system first, and your teams learn on the tool that exists.
Where the data is processed, whether it is kept, whether it trains a model: those are decided before the first line of code, not at acceptance testing.
How it works
Each step has a price, a timeframe and a deliverable. You decide on the next one from what the last one produced, not from a commitment signed at the outset.
€9,500 excl. VAT · fixed fee, one area of the business
Know exactly where generative AI would save you time, what it costs, and what to build first.
from €18,000 excl. VAT · fixed fee, one system in production
A first system genuinely running, used by your teams, with its gain measured.
on request · depending on what you take over
The system scales, it is monitored, and you take the controls at your own pace.
The levers
Each is treated as a subject in its own right, with its typical builds, its timeframe and its own data privacy question. And each one is backed by a build that happened.
Builds
A positioning audit, a content strategy, a rebuilt site and a CRM, then an assistant that reads the statistics and proposes the next moves itself.
The meeting transcript becomes a follow-up email and a priced proposal, sent after a quick read-through, while the CRM fills itself.
The application used to produce site selection studies, built with the analysts and used directly by them.
Data privacy
It is the first objection directors raise, and it is a fair one. It has precise technical answers, and which one applies depends on how sensitive your material is.
How I handle the questionFrom a programming interface under contract through to an open model running on a machine you control, depending on what you handle.
Where the data is processed, whether it is kept, whether it trains a model, which subprocessors see it, and for how long it is retained.
The architecture separates what may leave from what must not. Plenty of builds run on cut-down or anonymised data.
10+ years in the field
Data science & AI in production
Who you are dealing with
Ten years of data science inside large French companies, then AI systems put into production at software publishers, retailers and customer contact centres.
Since March 2026 I have run generative AI adoption projects at my main client, a 20-person consultancy, and at a remote monitoring company. The results shown on this site come from that work.
"What I learned there fits in one sentence: the obstacle is almost never the technology, it is knowing what to build first."
Guides
Positioning, content, a rebuilt site, a CRM, then an assistant wired into the acquisition data. The full account of one project, and what it produced.
A course almost always comes too early. What it cannot tell you about your own company, and the ten-day exercise to run before you book one.
The six questions to put to any supplier, a four-level sensitivity scale, what the GDPR actually requires, and a checklist to run before you go live.
Next step
Ten days to map where your teams lose time, put a figure on each lever and name the first system to build. If AI is not the right answer, I will tell you so.
Diagnostic
Étape 110 days
€9,500 excl. VAT · fixed fee, one area of the business