Contents
- Why product owners and project managers should understand agentic coding
- How the role actually changes
- Contents of a workshop for product owners and project managers
- Hands-on exercise: from user story to agent-ready specification
- Planning, estimation and steering in agentic projects
- Prerequisites, format and cost
- Conclusion
- Frequently Asked Questions
An agentic coding workshop for product owners and project managers explains how coding agents turn requirements into software – and what you need to do so that the right thing gets built. You learn to write stories and acceptance criteria that agents and developers can work with precisely, and how planning, estimation and acceptance change. No programming skills are required.
Why product owners and project managers should understand agentic coding
When development teams work with coding agents, the bottleneck moves. Writing code gets faster; clarifying requirements does not automatically follow. In practice, vague stories that an experienced developer used to "fix" by asking questions are now implemented literally and at high speed by an agent – gaps included. The result looks finished but misses the actual need.
For product owners and project managers, this means the quality of your specification affects the outcome more strongly and more quickly than before. At the same time, new opportunities appear: prototypes are available in hours instead of days, variants can be explored in parallel and feedback loops get shorter. If you understand how agents work, you can use these opportunities deliberately rather than just reacting to changed velocity numbers. Our article Agentic coding for non-developers provides a first overview.
How the role actually changes
The comparison below shows typical PO and PM tasks and how they change in teams using coding agents. In the workshop, it serves as a starting point for discussing your own processes.
| Task | Before | With coding agents |
|---|---|---|
| Writing user stories | Rough description, details clarified in conversation | Precise acceptance criteria, edge cases and examples up front |
| Prototyping | Click dummy from design, implementation later | Working prototypes early in the sprint, several variants |
| Estimation | Story points based on implementation effort | More weight on clarification, review and testing than on typing |
| Acceptance | At the end of the sprint | More frequent and granular, because results arrive faster |
| Risk management | Focus on schedule and budget | Also quality, security and maintainability of generated code |
Read more about this shift in How AI is changing the product owner role.
Contents of a workshop for product owners and project managers
A good workshop for this audience combines understanding, method and practice. Typical building blocks include:
- How coding agents work: plan, execute, verify – shown live and without jargon. You see what information an agent needs and where it typically takes a wrong turn.
- Agent-ready requirements: writing acceptance criteria, examples, non-goals and edge cases so they are unambiguous for both machines and people.
- Specifications as a steering tool: the spec-driven development approach makes requirements the central source of truth for agents and the team.
- Prototyping with agents: how to work with the team to produce testable variants quickly, so decisions are based on users rather than assumptions.
- Planning and process: what should change in sprint planning, the definition of done and review capacity.
- Limits and risks: data protection, security flaws, technical debt – and which questions you should ask your team.
Hands-on exercise: from user story to agent-ready specification
A core part of the workshop is working on real requirements from your backlog. A typical sequence looks like this:
- Pick a story: a medium-sized requirement your team plans to build soon, such as a filter feature or a new approval step.
- Find the gaps: together, work out which questions a developer would ask. What happens with empty data, missing permissions or very large lists?
- Sharpen acceptance criteria: write concrete, testable criteria, ideally with examples in a "Given – When – Then" format.
- Record non-goals: state explicitly what is not part of the story. Otherwise, agents tend to fill gaps with plausible but unwanted assumptions.
- Watch the implementation: the trainer has an agent implement the story. You see immediately which wording works and where misunderstandings arise.
- Accept the result: check whether the result meets the criteria – and which questions business acceptance still has to answer.
This exercise shows more clearly than any slide why precision in requirements becomes the most important lever.
Planning, estimation and steering in agentic projects
Project managers often ask in the workshop how to deal with fluctuating speed. Agents complete some tasks in a fraction of the previous time, while others take just as long because business clarification, integration or testing is the bottleneck. This leads to practical recommendations:
- Do not compare velocity after adoption with earlier values; let it settle anew.
- Plan review and testing capacity deliberately – it becomes the bottleneck more often than implementation.
- Measure lead time from "idea" to "in production" rather than only story points per sprint.
- Slice smaller increments so that fast implementation also leads to fast feedback.
How to put this into practice in Scrum or Kanban teams is covered in Integrating agentic coding into Scrum and Kanban.
Prerequisites, format and cost
Participants in an agentic coding workshop for product owners and project managers do not need programming skills. It helps to bring a few real requirements from your own backlog and, if possible, a developer or tech lead who joins the hands-on exercise. That way, business and engineering build a shared understanding right away.
Content, format (in-house or remote) and scope are agreed individually – for example as a compact introduction for the product team or combined with a training for developers. We place AI developers from real-world practice as trainers. Pricing is available on request by e-mail; workshops start from EUR 799 net.
Planning an app? Request a free fixed-price quote in 2 minutesConclusion
Coding agents make implementation faster, but not clarity. Product owners and project managers therefore become more important, not less: your requirements directly steer what gets built. A focused workshop shows how agents work, what agent-ready specifications look like and how planning and acceptance should adapt. Learning this early prevents quickly built wrong turns and channels the new speed into more user feedback.
Frequently Asked Questions
Do product owners need programming skills for an agentic coding workshop?
No. The workshop is designed for business roles and explains how coding agents work without programming exercises; the focus is on requirements, planning and acceptance.
Can product owners build prototypes with agents themselves?
For simple, clearly scoped prototypes this is possible in agreement with the team. Production code, however, should still be reviewed, tested and owned by developers.
Does estimating user stories change with coding agents?
Yes, because implementation often gets faster while clarification, review and testing do not speed up to the same degree. Teams should recalibrate their velocity and measure lead times.
How much does a workshop for product owners and project managers cost?
Cost depends on content, format and scope and is quoted on request by e-mail. Workshops start from EUR 799 net.