magiclabssenior AI engineering
Currently 1 analysis slot per month

AI Potential Analysis

2 weeks. Fixed price. A working prototype.

In 2 weeks you'll know exactly where AI makes your business money — and see the first use case already working.

Most companies know they should be “doing something with AI”. The expensive question is: what first? Most AI projects don't fail on the technology — they fail on the wrong priorities.

No workshops, no meeting marathons: a structured intake, your documents, two short calls. In between, I do the work — not you.

Request a 15-minute call€3,900 fixed price, plus VAT.

Why me

I don't just advise — I build production AI: assistants, RAG systems, agents and automations. Every day, as the operator behind several products. No agency overhead, no slideware.

See what we build

How it works

01
Week 1

Analysis

You fill in a structured intake and share documents: process descriptions, example files, your tool list. A short kickoff call (30 min, remote) clears open questions. Then I analyze your operations: Where does manual effort pile up? Where is data sitting idle? What does a mistake cost you?

02
Week 2

Proof of Concept

For the most promising use case I build a working prototype — with your real data and documents, not slides.

03
Wrap-up

Results session

Live demo of the prototype (60 min) and handover of the report — including a plain-language recommendation on how to proceed.

What you get

€3,900 · fixed price, plus VAT.

Prioritized AI roadmap

5–10 concrete use cases, scored by effort, value (€/month) and risk.

Working proof of concept

The top use case runs at the end of the two weeks — with your data.

Implementation plan with cost frame

Concrete enough to get started — with me or in-house.

Straight talk where AI isn't worth it

An honest assessment saves you more money than any slide deck.

Guarantee

If we don't find a use case with realistic ROI in week 1, we stop — and you pay only half.

For software teams

AI Developer Productivity Audit — same fixed price

For software companies and IT teams: a two-week audit of your development workflow. Where do AI-assisted development practices (Claude Code, agents, automated code reviews) save the most engineering hours? The tooling is set up directly in your repository and the effect is measured on a real task — your team gets an adoption playbook.

“Your developers get 30–50% faster — I'll show you how in 2 weeks. Measured on your own code.”

What happens after?

Optionally, I implement the first use case at a fixed price (from €7,500, production-ready in 2–6 weeks). Book the implementation within 30 days and 50% of the analysis fee is credited. For ongoing support there's a monthly-cancellable AI engineering retainer.

MK

Who does the work?

Done personally by the owner — CTO of a German AI SaaS company who ships LLM agents, RAG systems and automations to production every day. An operator, not an agency: senior execution without agency overhead, and no strategy workshops that end without working software.

Claude / AnthropicOpenAILLM AgentsRAGn8n / MakeTypeScript / Next.jsPostgres / Supabase

Mike Koene

15 minutes is enough to judge it.

no pitch, no slides — just an honest assessment

Request a call