01
A realistic enterprise stack
Git, ticketing, wiki, team chat, browser-based development environment with Claude Code as the development agent. The setup mirrors a typical corporate environment - not a playground, not a demo rig.
The development process of the future. To touch, to try, to rebuild - just not on your stack.
Once AI knows the entire development process - meeting notes, requirements, tickets, code, deployment - the pace of development shifts considerably. That part is real and it is measurable. But pace without the discipline to verify is not progress. It is production risk with a shorter lead time.
Both halves belong together, and neither one lands on a slide. Your team has to experience how fast an agent delivers - and, in the same run, how convincingly it can be wrong. So the real question is not whether AI-assisted development works, but under which controls it does. And how your team gets there without rebuilding its own stack, without an integration project, without months of approval loops.
That is exactly what the Validation Lab answers: a complete environment where your team does not watch the process of the future but runs it - analysing requirements, cutting tickets, steering the agent, reviewing results and deploying them. Without your stack, without your data, without risk. Afterwards you decide what to adopt - based on your own experience instead of demos and benchmarks.
- Format
- Fully remote
- Team size
- Up to 10, 8 recommended
- Environment
- Dedicated instance, EU hosting
- Price
- EUR 7,000 fixed price (Orientation)
- Availability
- Limited, ask early
What you get
A complete environment. Exclusively yours for the duration of the workshop.
Built for your workshop, removed without a trace afterwards. And inside it: no spectator seats. Everyone gets their own access and works hands-on from the first login.
02
Tool-neutral - transferable to your stack
Which products the lab happens to run is secondary. What your team learns is the process: how an agent gets its context, where verification has to bite, which handovers between roles hold up. Those patterns are the same whether your tracker is Jira or Azure DevOps and your wiki Confluence or SharePoint. What your team takes away fits practically any enterprise stack.
03
No customer code, no customer data
The work happens on a prepared project. Nothing to release, nothing to anonymise, no security review loop before the start. You could begin tomorrow.
04
EU hosting, dedicated instance
The environment runs on European infrastructure and belongs exclusively to you for the duration of the workshop.
05
Current models side by side
Work with models from different providers and find out what fits your workflow. All tokens are included in the workshop price (fair use).
For comparison: building such an environment yourself costs several person-weeks - planning, configuration, integration, testing. Here you arrive on day 1 and start working.
What your team takes away
Four things that outlast the workshop.
A realistic picture
Of what AI-assisted development delivers in your way of working - and where its limits are. No marketing slides, your own experience.
The full process in your own hands
From meeting notes and requirements through epics, stories and tickets to implementation by the agent - testable immediately. Which is why the workshop is not only for developers: project managers and agile coaches work hands-on from the first hour.
Verification as a core skill
AI is a team member of a peculiar kind - tirelessly fast, never contradicting, always convincing, even when it is wrong. That is why the training project deliberately contains the contradictions, gaps and outdated decisions of real project documentation. Exactly there your team learns where persuasiveness and correctness diverge - in a protected space instead of in production.
A solid basis for your adoption decision
Which workflows to adopt, which guardrails you need, where the investment pays off - framed for your stack, not the lab's.
I accompany your team through the entire workshop - as a coach at your side, not as a presenter at the slide deck.
Formats
Two paths, one goal.
Orientation
For teams at the start of AI adoption, with mixed levels of experience.
We start with a compact foundations block that brings everyone to a shared level - after that, the time belongs to practice. In pairs, your team runs through the complete development process on the prepared training project: analysing requirements, cutting tickets, implementing with the agent, reviewing results and deploying. All done by your team, nothing demonstrated. The learning moments are built into the project - they do not happen by accident, they are planned.
The result: your team comes out speaking a common language for LLM-assisted development - from the context window to the review gate. Every person has run the full process once themselves, not watched it. And everyone can judge realistically which parts of it fit their own working day and which do not.
Included in the fixed price: coaching, environment, tokens, preparation and follow-up. No hidden costs. For those two days I coach your team and no one else - which is why dates are scarce.
Validation
For teams that already have a concrete way of working in mind and want to put it to the test under realistic conditions.
I configure the environment to match your process - roles, workflows, gates, tooling. Your team validates and optimises it in live operation, with support throughout. Duration and scope follow your process.
The result: your process has been confirmed or disproved under realistic conditions. You know concretely which roles, gates and tooling decisions hold - and where to adjust before the process meets real projects.
You receive the fixed price after a short kickoff call - no surprises, no open ends.
Ground rules
What else you should know.
- Fully remote
- No travel overhead, distributed teams explicitly welcome.
- Up to 10 people, 8 recommended
- Intensive coaching instead of a mass webinar.
- Failure is a feature, not an incident
- The Validation Lab is a deliberately open learning environment - free of production constraints, with maximum room to experiment. Which is precisely why it is not meant for live projects: here, things are allowed to go wrong that must never go wrong in production.
- Dedicated setup and teardown
- Every instance is built specifically for you, can be adapted to your own processes, and is removed completely after the workshop.
- Limited availability
- For the Orientation format, personal coaching means exactly that: during those days my time belongs to your team, and I run no second workshop alongside it. Which leaves few dates in a year. Validation engagements schedule more flexibly - but there too: if you have a particular window in mind, it pays to talk early.
And afterwards?
From workshop to day-to-day operation.
The workshop delivers the insights - moving them into regular operation is the next step. If you want, I will guide you through it: which controls belong in the structure, which processes to change, and how AI-assisted development lands safely in your everyday work.
One experience nearly every team takes away from the lab: at some point the agent is convincingly wrong. Whether anyone notices must not depend on who happens to be reviewing, and on how disciplined that review is on that particular day. Verification carried by individuals does not survive deadline pressure. It belongs in the structure - in gates that hold even when nobody is looking.
That is what Dissemination Control stands for: deterministic, structural controls instead of probabilistic inspection. The blueprint is openly documented, vendor-neutral and free to use - with or without me.
Read the blueprint on GitHub ↗ The next step
Let us find out whether the Validation Lab fits your team.
Schedule a no-obligation consultation. I listen, tell you openly which format fits - and whether any of them does.
Schedule a consultation →P.S.: You are a developer, you are convinced, but you are not the decision-maker? Forward this page to your CTO - one conversation, no obligation.