Ceremonies

These are the recurring events that keep the process running. Each entry lists who attends, the timebox, and how AI supports it. Stakeholders are only required at the Sprint Review (and optionally Planning).

Quick reference

Ceremony	When	Timebox	Required attendees
Sprint Planning	Day 1	2–4 h	Whole team; stakeholders (PO) for priorities
Daily Scrum	Every day	15 min	Whole team
Backlog Refinement	Mid-sprint	1–2 h	BA, Architect, + reps; stakeholders (PO) consulted
Sprint Review	Last day	1–2 h	Whole team + stakeholders (PO)
Retrospective	Last day	1–1.5 h	Whole team (no stakeholders)

Sprint Planning

Purpose: agree the Sprint Goal and select the Sprint Backlog. Inputs: prioritized, “Ready” product backlog; team capacity; previous velocity. Outputs: Sprint Goal, committed Sprint Backlog, task breakdown.

AI support: AI proposes story-point estimates from historical similar items, flags items that look under-refined, and forecasts a completion probability for the proposed scope. The team treats this as a second opinion, not a decision.

Daily Scrum

Purpose: synchronize, plan the next 24 hours, surface blockers. Inputs: the board state; yesterday’s progress. Outputs: an updated plan; a list of blockers for the PM/Scrum Master to clear.

AI support: an automated stand-up digest pulled from the board, commits, and PR activity is posted before the meeting, so the 15 minutes focus on blockers and decisions rather than status recitation.

Purpose: make upcoming items “Ready” - clear, estimated, with acceptance criteria. Inputs: roadmap, new requests, stakeholder feedback, tech-debt register. Outputs: refined, estimated, prioritized backlog items meeting the Definition of Ready.

AI support: the BA uses AI to draft user stories and acceptance criteria from raw notes; the Architect uses AI to sketch design options and surface risks. Humans edit and approve.

Sprint Review

Purpose: demo the increment, gather stakeholder feedback, adapt the backlog. Inputs: the “Done” increment; the Sprint Goal. Outputs: stakeholder feedback, updated backlog priorities, release notes.

AI support: AI drafts release notes and a plain-language demo summary from merged work, so the team walks in with a polished narrative. This is the one ceremony where stakeholders (acting as Product Owner) are essential - they give feedback and accept or reject the increment.

Retrospective

Purpose: improve how the team works. Inputs: sprint metrics (velocity, cycle time, defect counts, AI-assist stats), team sentiment. Outputs: 1–3 concrete, owned improvement actions for next sprint.

AI support: AI clusters qualitative input and correlates it with metrics to surface themes (e.g. “review wait time rose 40% this sprint”). The team decides what to act on.

flowchart LR
    M[Sprint metrics] --> AI[AI clusters themes]
    F[Team feedback] --> AI
    AI --> D[Team discusses]
    D --> ACT[1-3 owned actions]
    ACT --> NEXT[Next sprint]