PromptPartner DE

Test first, then build.

We test one job on cases you've already finished. If it works, we set up the agent and look after it every month.

Six colleagues look at one laptop together
Your decisions, step by stepIllustration
  1. Trial runYour cases side by side
  2. AI assessmentFixed-price proposal or a no
  3. Set-upAcceptance on past cases
  4. Every monthcarry on or stop

You decide after each step

Where most firms start: one job

Licences bought, training done, and the back office still works by hand. That's where we start, with one job. Firms with 10 to 250 staff start with the trial run. Larger firms start with the AI assessment.

  • One job always waits on the same personWe run an agent, read only, over cases you've already finished, and its draft sits beside your original. See the trial run

  • You have to bring in AI, but nobody leads it yetA fixed-term AI lead takes the questions now and hands over to someone on your staff later.

  • A larger firm with several departmentsHere the AI assessment comes first, sized for several departments. See the AI assessment

If it works, this is how it goes on

You agree each step on its own, and you can stop after any of them.

  1. 01

    Trial run

    An agent's drafts beside cases you've already finished, then a clear yes or no.

  2. 02

    AI assessment

    The AI Value Canvas with your people, ending in a fixed-price proposal or a no.

  3. 03

    Set-up

    We set up the agent and test it on your past cases.

  4. 04

    Looked after

    Every month: new price lists, changed rules, a check against fixed examples and a one-page report.

Once the first agent runs, the next one can build on the same data.

At the end, an agent does one job for you

A chat answers one question for one person. An agent does a whole job. It takes data from several of your systems, applies your rules and hands the result to a person to check.

  • ERP
  • Mailbox
  • Files
  • CRM

The agent

Works by your rules

skillsplugins

Draft: quoteExample

Your team approves itNothing goes out without a yes

Set up before the first run

  • AccessOnly the folders and systems you allow.
  • GuardrailsRules the agent works within.
  • LogEvery step written down.
The agent works between your systems. One of your people approves what goes out.

Unless you choose otherwise, your agents run in your own environment and we operate them. Microsoft 365 is an important place for agents. It shouldn't be the only one, so we keep a second way open: your own server and open models.

Five things we put in writing

  • DataAt first we only read. Before the first file comes to us, we put in writing where it is stored. What we measure belongs to you.

  • Plain answerIf your software can already do the job, or your data can't carry it, we tell you and stop there.

  • Fixed priceFrom the AI assessment on, each step has a fixed price and a fixed scope, and we bill no hours.

  • StopYou can stop after any step.

  • AcceptanceThe monthly fee starts only once the agent passes acceptance on your own past cases.

Questions about how we work

What does it cost?

From the AI assessment on, each step has a fixed price. It depends on the size of your firm and on whether it covers one job or several. We name it in the first call.

Can we stop after the AI assessment?

Yes. After the AI assessment you decide, with your own numbers, whether to go on. What we measured belongs to you.

Do we have to do every step?

No. You can stay with one job for as long as it suits you. We suggest the next step only once the first one is running.

Who runs the agent after the set-up?

We do, unless you choose otherwise. We look after it every month and send you a one-page report. If you stop, the agents, the rules and the data stay with you.

Which job should we try first?

30 minutes with Lukas. You leave with at least one recommendation you can use, whether or not we ever work together.