The A3 Delegation System Pilot Workshop

Turn an AI Workflow Into an Explicit Delegation

Your team already delegates work to AI: reports, research, customer feedback analysis, stakeholder communication, or parts of operational workflows.

But can you answer these questions without improvising?

• What may AI decide, and what must remain a human decision?
• What does "good enough" mean for this particular work?
• Who verifies the result before somebody acts on it?
• Who checks whether the delegation still works after the model or workflow changes?

If those answers live in one person's head, or nowhere, your problem is no longer prompting. You have a delegation problem.

The A3 Delegation System gives you a practical way to decide what AI may do, hand over the work clearly, define acceptable results, and inspect the AI delegation over time. The class is in English. 🇬🇧

VAT Included for Customers Situated in Germany

Prices include 19% VAT for buyers situated in Germany.

Electronic Withdrawal Function for Consumers in the European Union

If you concluded this contract online as a consumer situated in the European Union and have a statutory right of withdrawal, you may submit your withdrawal statement through our electronic withdrawal function.

Bookings by companies, public authorities, organizations, self-employed professionals, freelancers, or other business customers are governed by our Terms and Conditions, in particular the provisions on withdrawal from participation, cancellation, rebooking, substitute participation, and refunds.

Before You Book: This Is the 2nd Public Workshop

This is the second pilot workshop, not a polished training product with an established public track record.

The underlying method and working artifacts exist. The public workshop format is in its second iteration. Explanations may need revision. Exercises may take longer than planned. An artifact that looks clear to its creator may fail when another participant tries to use it.

That uncertainty is why the price is $249.

You are not receiving a conventional discount on a finished product. You are accepting some of the risk of an early format. In return, you receive the complete working method, direct coaching, and the opportunity to influence the next version through a candid closing retrospective.

Do not book if you want a predictable, polished course supported by testimonials from previous A3 Delegation System public training classes. Book if you want to test a serious working method while it is still being shaped, and you are willing to let the method expose gaps in your own AI workflow.

What You Will Do

All three sessions work through one prepared case: Brandtner Pumps, a family-owned manufacturer whose project lead wants an AI assistant to draft the monthly status report to the board. You follow this delegation from the first decision to the month the AI-drafted report goes wrong, so you practice every step on realistic material without bringing your company's data into the room.

1. Decide: Assist, Automate, or Avoid

Classify what role AI should play:

• Assist: AI drafts; a human reviews and decides.
• Automate: AI executes within explicit rules and inspection boundaries.
• Avoid: The work remains human because delegation is unjustified or too risky, given the involvement of trust and relationships.

The objective is not maximum automation. It is an explicit autonomy decision that someone else can understand.

2. Hand Over the Work

Turn "use AI for this" into an A3 Handoff Canvas that clarifies purpose, inputs, outputs, responsibilities, boundaries, records, and stop conditions.

3. Define Done

Replace "a human checks it" with task-specific acceptance criteria. Decide what must be verified, by whom, against which examples, and before which downstream action.

4. Inspect the Delegation

Define a delegation owner and review rhythm for detecting output drift, changed assumptions, or the quiet shift from supervised assistance to unsupervised automation.

What You Will Leave With

During the pilot, you will help Marta of Brandtner Pumps to fix her board meeting status report challenge:

• An A3 classification explaining the appropriate level of autonomy.
• An A3 Handoff Canvas documenting the delegation.
• An AI Definition of Done with task-specific acceptance criteria.
• A Delegation Audit with an owner, review rhythm, and first inspection point.

These artifacts are not certificates that a delegation is complete, compliant, or safe. Their purpose is to expose missing decisions while you can still change them.

Workshop Schedule

Session 1: Decide, Hand Over, and Define Done

Thursday, December 3, 2026 · 2:00–6:00 pm CET

• Classify agile tasks as Assist, Automate, or Avoid, and learn to spot the ones that sit near a boundary.
• Turn the project lead's plan into an A3 Handoff Canvas: task split, inputs, outputs, validation, failure response, and records.
• Decide which model tier is good enough for the task, and why an over-powered model fails the test, too.
• Write an AI Definition of Done for output that is never the same twice, using a report written by hand as the reference case.

Session 2: Inspect and Adapt

Thursday, December 10, 2026 · 2:00–6:00 pm CET

• Recap session 1, so the whole system is fresh before the case jumps five months ahead.
• Audit a delegation that has quietly drifted, with four Check Cards, a verdict for each, and the evidence from the case.
• Record your decisions in the Re-classification Log and define what must be true before the AI may draft again.
• Carry those decisions back into the Handoff Canvas and the AI Definition of Done, closing the loop through all six stages of the A3 Delegation Life Cycle.
• See how the Roll-up turns audit results into the cost and value evidence your CFO and governance people will ask for.

Session 3: Take Stock and Agree

Thursday, December 17, 2026 · 2:00–6:00 pm CET

• Find every place AI already touches the team's work, including the shortcuts nobody approved, and record them in an AI Workflow Inventory.
• Group the inventory into task classes that share a standard, a reviewer, and a model tier.
• Write the AI Working Agreement: the one page a new colleague should read on day one, with standing defaults, hard boundaries, and a way for anyone to pause a workflow they do not trust.
• Walk through the complete system one last time and choose the first step you will take with your own team.
• Retrospective on the pilot workshop itself.

This Workshop Is for You If…

• You already use AI in recurring professional work, but cannot clearly explain the operating rules behind it.
• You brief an internal team or external partner on work whose output can vary.
• You are a Product Owner, Product Manager, Scrum Master, Agile Coach, Project Manager, team lead, consultant, trainer, or operations professional helping a team adopt AI.
• You want a practical alternative to both uncontrolled experimentation and compliance-first bureaucracy.
• You are willing to bring one real workflow and let peers challenge your assumptions.
• You want to be able to introduce disciplined AI delegation to a team or organization.

No coding is required. This is not an introduction to ChatGPT, a prompt-engineering class, a tool comparison, or a workshop for building an AI agent. The focus is on the human work of deciding, briefing, accepting, and inspecting delegated work.

Practical Details

• Session 1: December 3, 2026, 2:00–6:00 pm CET
• Session 2: December 10, 2026, 2:00–6:00 pm CET
• Session 3: December 17, 2026, 2:00–6:00 pm CET
• Price: $249 pilot-workshop price
• Participation: The workshop depends on live exercises, group work, and candid feedback.

Your Trainer

Stefan Wolpers has worked with agile teams and organizations for more than two decades. He is a Professional Scrum Trainer, author of the Scrum Anti‑Patterns Guide (Pearson’s Professional Scrum Series), and creator of the A3 Delegation System.

This workshop grew out of a practical problem: teams are getting better at producing work with AI faster than they are at explaining, evaluating, and maintaining the delegations behind that work.

Expect direct feedback, working sessions rather than extended lectures, and open discussion of where the workshop itself still needs improvement. Learn more about Stefan.

Berlin Product People GmbH
Stöckter Deich 101
D-21423 Winsen (Luhe)

Email: info@berlin-product-people.com
Website: https://berlin-product-people.com

Amtsgericht Lüneburg, HRB 212146
Geschäftsführung: Stefan Wolpers