Your daily scrum runs long. Your backlog is a junk drawer. Your retro produces the same three sticky notes every sprint. You shouldn’t have to wait for a coach to be available to get unstuck — so I built one that’s always on.
What to ask it
- Sprint planning — capacity, complexity, dependencies, and what to leave out
- Daily scrum — keeping it to 15 minutes and surfacing real blockers
- Backlog health — refinement, acceptance criteria, and a backlog that’s ready
- Retrospectives — turning “we should improve that” into an owned action
- Dependencies & flow — blockers, aging work, predictability, and PI health
- AI + Jira — using AI as the accelerator with Jira as the system of record
It answers the way I coach: direct, practical, two to four things you can do this sprint — and it points you to the right playbook on this site when there’s more to read.
Built responsibly
I advise on AI governance, so I built the Scrum Agent the way I’d want a client to:
- It says it’s AI. It’s an assistant, clearly labeled — never passed off as me.
- It knows its lane. It sticks to Agile and delivery and steers anything else back.
- It doesn’t want your secrets. Please don’t paste confidential company or client data — the same rule I hold every team to.
- It has fair-use limits. Daily limits keep it available for everyone.
- It knows when to hand off. When the problem is bigger than a chat, it tells you so.
AI answers can be wrong. Treat it as a fast first answer — the coach, the team, and the context decide what actually happens.
Tell me how it did
Did it nail your question — or miss? I read every note. Send feedback, wrong answers, and questions you’d like it to handle better to wendell@agilebykillette.com.
When it’s bigger than a chat
AI should reduce the work around delivery — not replace the thinking behind it.
Stalled transformation, struggling train, or AI rollout that needs governance? Book a conversation with Wendell →