Internal test
If the plan is simple enough, your team can run the experiment internally.
AI Workflow Diagnostic · €2,500
Turn one promising AI workflow opportunity into a clear experiment plan.
For B2B SaaS and platform teams that already know which workflow, productization question, or operating-model decision is worth exploring, but need clarity on owners, metrics, risks, human checkpoints, data needs, and implementation options before building.
Opens a booking calendar. No payment required to start the conversation.
€2,500 · One week · 2 to 3 working sessions · One workflow opportunity · Experiment plan · No production access required
When to start here
The diagnostic is the right next step when the workflow opportunity is already visible, but the experiment is not yet clear enough to run.
Good fit
- You already know which workflow needs attention.
- The team agrees the problem matters.
- The next step requires more than a 90-minute review.
- You need to clarify owners, data, systems, risk, and success metrics.
- You want a practical experiment plan before committing to a build or sprint.
Not a good fit
- You are still unsure which workflow to start with.
- You mainly want AI inspiration or training.
- You want to automate without inspecting the current workflow.
- You need broad AI strategy detached from a product, workflow, or operating-model question.
What we work through
- Current workflow and decision points
- User, customer, or team pain
- AI role and boundaries
- Data, tools, and context required
- Human checkpoints and escalation paths
- Quality, trust, and risk controls
- Success metrics and experiment design
- Implementation options and next step
What you get
You receive an AI Workflow Experiment Plan.
- Problem framing
- Workflow opportunity
- AI role definition
- Human control model
- Data and context requirements
- Risks and assumptions
- Success metrics
- Experiment design
- Implementation options
- Recommended next step
Where it can lead
Operator / Advisory Sprint
If the experiment needs hands-on help, we can continue into a four-week sprint.
Stop or delay
If the workflow is not ready, the recommendation can be to stop, delay, or choose a different starting point.
A few things people ask
What actually happens during the week?
We run 2 to 3 working sessions with the people who own the workflow. Between sessions I reconstruct the current workflow, test where AI could realistically help, and draft the experiment plan so each session reviews something concrete instead of starting from a blank page.
Do you need access to our systems or customer data?
No production access is required. Redacted, anonymized, or representative examples are enough. If your team wants to share more, that is your call, not a condition of the work.
What do we end up with?
An AI Workflow Experiment Plan. It covers the problem framing, the AI role, the human control model, data and context needs, risks, success metrics, the experiment design, and the implementation options, so the next decision has an owner and a measure.
How is this different from the workshop?
The workshop decides whether one workflow is worth changing at all. The diagnostic assumes that question is settled and goes deeper on owners, metrics, risk, and how the experiment would actually run.
Know the workflow, but not the experiment?
The diagnostic turns one promising AI workflow opportunity into a clear experiment plan with owners, metrics, risks, human checkpoints, and implementation options.
€2,500 · One week · 2 to 3 working sessions