The standard assessment
Interviews and a workshop, ending in a slide deck of what people said they do. An opinion survey with a big invoice attached.
You get impressions.
GreenPT Advisory
You can't govern what you've never measured.
AI is already running inside your organisation. In tools nobody registered, on budgets nobody added up, with data nobody mapped. We start by measuring it: what it costs, where the data goes, what it emits, and who's using what.
What we find
Four numbers your organisation can't produce today.
In four weeks, we replace every guess below with a figure you can defend.
Just one question
How is AI used in your organisation, and what is it costing you in money, risk, data and emissions?
Deliberately narrow, because that's where the answers are missing and where the exposure sits.
-
What you're actually spending
Across every team and every personal subscription, including what runs outside the IT budget.
-
Where your data is going
Which vendors, in which countries, and whether your data processing agreements actually cover them.
-
What you're emitting
The energy and carbon behind your AI use, none of which is in your reporting today.
-
How exposed you are
How much shadow AI is running that was never approved, and how dependent you've quietly become on a handful of providers.
Most organisations can't answer one of these. We hand you all four.
Why now: three forces are turning this into a deadline.
The three forces
- EU AI Act Regulation
Obligations are phasing in for organisations that use AI, not just those that build it. Knowing what you run is the first requirement, not the last.
- CSRD Reporting
The digital footprint is being pulled into scope, with procurement and accountability rules close behind for public bodies. "We don't know what our AI emits" is not an answer you can file.
- Unmapped Budget
AI spend is scattered across teams and cards. It rarely shows up as one line, which is exactly why it goes unquestioned.
This has been measured across thousands of organisations
The evidence
47%
of workplace AI runs on personal accounts, not company ones
Netskope, 2026 +267%
one-year growth in AI paid on personal cards and expensed back
Zylo, 2026 15%
of company software spend is still bought through IT
Zylo, 2026 64%
of AI vendors don’t name their AI subprocessor in the contract
DataGrail, 2026And yours? That’s what we measure.
The entry point
A gap analysis, not a discovery phase.
Four weeks. No commitment to anything after it. We measure what your organisation actually does with AI, and hold it against where you need to be on cost, ethics, privacy and emissions.
Not an infrastructure audit, not a digital transformation programme, not a forty-slide deck about your future.
What we measure
- Which AI tools are in use, including the ones nobody registered.
- What you're spending, in total and per team.
- Where your data flows, and under whose jurisdiction.
- What your AI use emits.
What you get
- A live dashboard of AI use, spend and emissions, which stays yours after we leave.
- A ranked list of gaps, from "fix this quarter" to "watch this".
- A recommendation you can take to your board, your DPO or your procurement team.
Why us
Most consultants ask what you do. We measure it.
GreenPT Advisory
We build the platform, so we read real usage, real tokens, real spend and real emissions. When the four weeks are over, the dashboard stays live.
You get an instrument.
If you need more
Insight first. Implementation only if you want it.
You decide after the analysis, with the findings in front of you. Not before.
- Weeks 1 to 4
The gap analysis
Measure, rank the gaps, recommend. You end with a dashboard and a clear picture. You are here.
- The next 12 weeks
The implementation programme
From insight to control: an AI policy people will follow, supplier requirements you can drop into a tender, training, and monitoring that outlasts the project.
The fair question
You sell an AI platform. Doesn't that make you biased?
It's the first thing we'd ask, so here it is up front. Yes, we build GreenPT. And no, the analysis doesn't end in a pitch. We measure your situation, not our fit.
The dashboard, the gaps and the supplier requirements are yours to use with any vendor you like. Where we think GreenPT is the right answer we'll say so, with the numbers next to it, and you're free to hold those against anyone else's. We'd rather be the advisor you keep than the vendor you regret.
Frequently asked questions
Answers, grouped by who is asking.
Every question carries a link, so you can forward a single answer to the department that needs it.
What exactly do I get after four weeks?
A dashboard showing your actual AI usage, spend, data routes and emissions. A list of the gaps we found, ordered by how quickly they can be closed. And a recommendation you can take to your board, your privacy lead or your procurement team. The dashboard stays yours, even if you never buy anything from us again.
What does it cost, and what does the next step cost?
The gap analysis is a fixed-fee, four-week engagement. The optional 12-week implementation programme is a separate engagement with its own quote.
Are we locked in?
No. The dashboard and the underlying data are yours. There is no technical dependency that forces you back to us, and the implementation programme is a separate engagement you commission separately.
Can this be commissioned without a tender?
In most cases yes, because the engagement is a standalone piece of work with its own deliverable and no follow-on obligation. Thresholds and rules differ by country and by organisation, so check your own procurement policy. We are happy to supply whatever documentation your process requires.
What access do you need?
An export of twelve months of procurement and card transactions. Read-only access to your SSO logs, or an export of them. Four interviews. That is all.
Who actually does the work?
Robert Keus (CEO and founder, 15+ years in privacy-first digital work) leads the first week of every engagement himself. After that the measurement is carried out by our own team of AI engineers, the same people who build and run the models and infrastructure behind GreenPT. They spend their days turning raw usage into spend, emissions and data routes, so they know where AI hides in an organisation and how to quantify it. You get a founder's judgement on what matters, and a team that measures this for a living.
What happens to our data afterwards?
Data is stored on our EU-hosted, ISO 27001 infrastructure for the duration of the engagement plus 30 days for corrections, then deleted. Aggregated, non-identifiable measurements may be kept for our own baseline research, and we describe how on the methodology page. Placeholder: confirm the exact retention window before this page goes live.
Do you provide a data processing agreement?
Yes, before any data changes hands, along with a privacy impact assessment annex where your jurisdiction requires one.
Where is the data processed?
Entirely on European infrastructure: Scaleway data centres in France, powered by renewable energy, ISO 27001 certified. Named subprocessors are listed in the DPA. Placeholder: confirm the exact list before this page goes live, as this is where we structurally win against vendors who do not name theirs.
You sell an AI platform. Doesn't that make you biased?
It's the first thing we'd ask, so here it is up front. Yes, we build GreenPT. And no, the analysis doesn't end in a pitch. We measure your situation, not our fit. The dashboard, the gaps and the supplier requirements are yours to use with any vendor you like. Where we think GreenPT is the right answer we'll say so, with the numbers next to it, and you're free to hold those against anyone else's. We'd rather be the advisor you keep than the vendor you regret.
Ready to start?
Find out where you stand, before someone asks you to explain it.
- Four weeks
- Fixed price
- A dashboard you keep
30-minute intake. No presentation, no further commitment.
DPA and privacy annex supplied up front.
Robert Keus Founder, GreenPT
Call before you request anything. Thirty minutes, no presentation. I'll tell you what we found at comparable organisations and whether this is worth doing for you.