Service ladder — Step 2

Put the right tools in place — and build where it pays off.

Set up ChatGPT, Claude or Copilot properly for your team. And build AI agents where they genuinely save time. No generic rollout — what works for your processes and your people.

Book a meeting →

Software-independent · Quick wins first · SMEs

Deliverables

What you concretely get.

Tools set up properly

ChatGPT, Claude, Copilot or another tool — we set it up for your context. System prompts, workflows and integrations that match how you actually work, not a generic demo.

Building AI agents

Where an agent delivers more than a tool, we build it. Lead follow-up, email processing, reporting — concrete automation in the places where your team loses time today.

Integrations & connections

Connected to the systems you already use — CRM, calendar, email. So AI isn't a standalone tool, but part of the way your team already works.

How it works

Quick wins first, then expand.

1

Choosing what pays off first
Based on the AI Startscan (or a short inventory), we determine which tools and agents have the most impact. Not everything at once — the quick wins first, so your team sees results straight away.
2

Set up & build
Configuring tools, building agents and connecting them to the systems you already use. I do it myself — no chain of suppliers, no waiting times.
3

Live, tested and handed over
We test thoroughly, the team gets to know what's in place, and everything is explained to you so you're not dependent on us for day-to-day use.

Who it's for

Does this sound like you?

You've done the AI Startscan and now want to actually set up the quick wins.
You want AI tools set up properly for your team — but don't want to figure out how everything should be configured yourself.
You want an agent that takes over a specific task — but from someone who actually builds it, not just advises.

Frequently asked questions

Answers to your questions.

Do you also work with the team that built the system?

Yes. If an external party or internal team built the system, I get familiar with the code and documentation and take over the guidance. I work together with whoever's already involved, not around them.

How long does an implementation project take on average?

The go-live phase typically takes 2–4 weeks. The optimisation cycle afterwards is ongoing — at least 3 months to get real trend data. After that you can choose a long-term retainer or an occasional check-in.

What if the system doesn't perform as expected after go-live?

Then we activate the fallback plan we drew up in advance, identify the cause and fix it. I take this seriously — a failed go-live is not an option if you do the preparation properly.

Ready to really set up AI?

Book a 20-minute meeting. Or start with the AI Startscan if you first want to know where the biggest gains are.

Book a meeting → Start with the AI Startscan →