The board wants AI. Nobody said what for.
Someone was given the task, with no method and no place to start.
You bought AI. Nothing has changed?
We give your team superpowers.
You know your team should use AI. The hard part is knowing where to start. We show you where, build it in your systems, and teach your people to use it.
For companies in Europe with €50 to €500 million in revenue or ARR, often founder-led or owned by private equity or a family office.
30 minutes with Lukas. You leave with at least one recommendation you can use, whether or not we ever work together.
Your rollout partner for 
AI that does the work
Companies we’ve built, scaled, or advised
Your AI world today
There is a line for AI in your profit and loss. There is no line for what it produced.
Licences bought, trainings done.
After one or two sessions, most people worked the old way again.
Your people use AI anyway.
Often on personal accounts. Company data can go in, with no contract and no log.
What shadow AI is, and how to bring it under controlEvery quarter like this, you pay for the licences, the personal accounts multiply, and the board asks again.
What helps is clarity: one or two use cases that pay off, and a company account people want to use. Our team builds them with yours.
Proof
Work that went live.
We work with mid-sized and larger companies in Europe, in any industry, from about €50 million to €500 million in revenue or ARR. Many are in software, SaaS or consulting, often founder-led or owned by private equity or a family office.
Three projects we completed recently. Anonymised where a client has not cleared naming.
€2.07M–€3.0M
value created a year, floor to all-in
A European software company above €100M ARR
We shipped 42 working skills across 15+ plugins. One of them turns a week of collateral work into minutes. With IT and Security, we also built the connecting layer and the governance for the Claude Enterprise rollout.
11 discovery sessions · 36 opportunities scored
30+
deliverables, all built and running
A 50M+-user European software company
We rebuilt its go-to-market AI-first: strategy, marketing, partner programme, sales and operations.
Weekly
senior-buyer meetings booked, running in production
A €250M+ European consulting firm
An AI outbound engine on ChatGPT Enterprise, connected to Salesforce and external B2B databases. Outreach follows buying signals and intent, written for each contact. Every BDR gets a call list each day.
Our engineers have also built data and AI systems in manufacturing, energy, transport, finance, healthcare, the public sector, real estate and security.
We also give back.
Two tools from our client work are open source under the MIT licence. Any team can use them.
Why PromptPartner
What changes for your team.
For the team you already have
The work nobody should do goes to AI.
Your people keep the work only people can do. Research before every call, the Monday report, the copy and paste: we hand that to AI.
Training on real work
Your people learn on their own tasks.
Every week while we build, your team brings real work to office hours and leaves with a better way to do it.
- 1Explore a useful practice
- 2Work on a team member’s task
- 3Share it with everyone
- 4Grow champions in every team
Where to start
A ranked list, with a euro figure on every use case.
The four-week assessment ranks your use cases and puts a euro figure on each, with the maths visible. It also says what to leave alone for now.
How the assessment worksSafe from day one
Your IT and legal teams set the rules with us.
They are right to worry. So they decide at the start, and everything we build follows those rules. Your data stays where it is.
- Control data access. Read-only first, and you decide what each workflow can reach.
- Keep human oversight. A person checks the work before anything is acted on.
- Build on your company’s rules. IT, legal and brand set them once, at the start.
- Move personal AI use to company accounts. People get a company account first. Then, on Claude or ChatGPT Enterprise, IT can block personal accounts on the company network.
Built in your systems
We build what we recommend.
An engineer builds the AI inside the tools your team already uses: the CRM, email and your documents. Then it goes live.
What we buildClaude or Codex skill
One request. A useful workflow. A person in control.
A skill is a recipe for work that comes back every week. It collects what it needs from your tools, prepares the work, and stops for your colleague to check it. Click through three examples.
Claude skillDemo · your systems connected: CRM, ERP, email
00:00 / 00:10
How it works
First we look. Then we build. Then your team runs it.
Four weeks · the assessment
We find where AI pays off.
We sit with your team, find the work AI should take over, and put a euro figure on each use case. We also list the AI tools your people already use, and what company data they put into them. You get a working prototype on your real systems.
You can stop here.Then · you choose
Our programme, or a project built for you.
Our ready-made programme: Claude or ChatGPT Enterprise for one department or the whole company, with use cases and training. We set it up together with your IT and legal teams. Or we build a custom AI project for you, also on local AI.
Ongoing · your team runs it
It keeps improving.
Your team runs it. We keep it running and add the next use cases, team by team.
Our method · Assess · Secure · Launch · Scale · Own
The assessment has a fixed price, set by the size of your company and by whether it covers one department or several.
The honest comparison
Why not do it yourself, or hire a consultancy?
Most companies try one of these first. Each has its strengths.
| Question | Waiting | A strategy consultancy | Doing it yourself | PromptPartner |
|---|---|---|---|---|
| Does something go live? | Waiting: No | A strategy consultancy: Often slides that nobody builds | Doing it yourself: Sometimes, when your power users find the time | PromptPartner: Yes, in your systems |
| Who trains your people? | Waiting: Nobody | A strategy consultancy: A workshop | Doing it yourself: Your power users, on top of their job | PromptPartner: We do, every week while we build |
| Are IT and legal on board? | Waiting: People paste data into any tool | A strategy consultancy: A policy paper | Doing it yourself: It depends | PromptPartner: They set the rules on day one |
| Do you know the value first? | Waiting: No | A strategy consultancy: A business case | Doing it yourself: Rarely | PromptPartner: A euro figure per use case |
| Who keeps control? | Waiting: Nobody decides | A strategy consultancy: The consultancy, until it leaves | Doing it yourself: You do | PromptPartner: You do: it is built in your systems |
Change begins with your people
A new tool is easy. A new habit takes partnership.
Some people are curious. Others are unsure where to begin. We give them a place to ask questions and to try AI on their real work.
Make it understandable.
We start with the work people know: what AI can do, where it needs checking, and how people stay in control.
Make it useful.
We take one task that comes back every week and work through it together. A small win people can feel starts the next conversation.
Make it part of the culture.
Shared learning, useful experiments, mistakes talked about openly. Your teams help shape their new routines.
Who you work with
The people who sit with your team.
We work with a European network of engineers and specialists. Two roles sit with your team on every project. The others join when a project is larger or custom.
“The value was never in the tool. It was in how the work ran.”
I spent twenty years building and selling automation software: first for web hosting, then for the cloud, then platforms sold through partners around the world, from €600K to €240M ARR. I had consultancies deliver 750 slides. We used nothing. So we build instead.
I take the first call myself.
€1.5B exit to CVC15 acquisitions since 20032.5 years deep in AI
Your core team
Forward-deployed engineer
Works inside your company and builds the AI into the tools your team already uses.
Business engineer
Maps how the work runs today, chooses the use cases with you and trains your people every week.
Extended team for larger or custom projects
Machine learning engineer
Models, tests and evaluations.
Data engineer
Connects your systems and data.
Software engineer
Integrations and dashboards.
Questions
What buyers ask us first.
What does PromptPartner do?
We deploy AI in mid-sized and larger companies. We start with a four-week assessment that finds the use cases that pay off, puts a euro figure on each one and builds one of them as a working prototype. Then the use cases go live, team by team, and your people learn to use them.
How much does it cost?
The four-week assessment has a fixed price, set by the size of your company and by whether the assessment covers one department or several. It is never billed by the hour. What comes after is a separate decision, made once you have the assessment's numbers.
Is our data safe?
Legal, IT and your brand team set the rules once. Your data stays where it is. Agents start with read-only access, and you control what they can reach.
Which AI tools do you work with?
We lead with Claude Enterprise and ChatGPT Enterprise, with the rules set once. You keep the licence contracts. We advise.
We already rolled out Copilot or ChatGPT. Where do we start?
With the work your teams repeat every week. The four-week assessment starts from the tools you already have. It finds the use cases worth doing and puts a euro figure on each.
Our people use AI on personal accounts. What do we do?
Give them a company account first. On Claude Enterprise and ChatGPT Enterprise, IT can then block personal accounts on the company network. The assessment lists the AI tools already in use. Our guide: promptpartner.ai/shadow-ai/
Four weeks from now, you’ll know exactly where AI pays off in your company.
What the 30-minute call gives you:
- An honest look at where AI could help your team first.
- At least one recommendation you can use, even if we never work together.
- A clear next step if it fits: a deeper call, then the assessment.
No pitch, no preparation needed. I look forward to it. Lukas