R&I task
In teams that use AI agents, part of the work can be produced very quickly: code, tests, documentation, prototypes, or analyses. But that output is not done until a person reviews it, validates it, and integrates it into the real system.
The R&I task exists to make that work visible.
What it includes
An R&I task may include:
- reviewing AI-generated code;
- checking that it meets a spec;
- running and analyzing tests;
- detecting architectural errors;
- removing duplication or technical debt;
- reviewing security, permissions, and sensitive data;
- integrating outputs from multiple agents;
- resolving conflicts between generated code and existing code;
- updating documentation;
- deciding which parts of the output are accepted, corrected, or discarded.
Why it matters
The usual risk is planning only the AI generation and forgetting the human review. The sprint looks feasible because the agent can generate a great deal in a short time, but it later collapses when everything needs review, validation, and integration.
In three-level estimation, the R&I task corresponds to the third level: human work dedicated to reviewing and integrating what AI has produced.
| Level | Type of work | How it is estimated |
|---|---|---|
| Zero-Point | Work carried out by AI. | Computational cost and minimal review. |
| Standard | Human work involving judgement, creativity, or decision-making. | Usual estimation techniques. |
| R&I | Review and integration of AI outputs. | Explicit time, usually in hours. |
Relationship with Scrum
During Sprint Planning, R&I tasks must appear in the Sprint Backlog. If they are not planned, the team implicitly assumes that reviewing is free. It isn't.
They are also connected to the Definition of Done. An increment containing AI output is not done until it passes human review and the checks defined by the team.
Best practices
- Estimate R&I before accepting AI-generated work into the sprint.
- Separate generation and integration into different tasks.
- Assign clear human owners.
- Review the highest-risk changes first.
- Record which output is accepted, modified, or discarded.
- Measure the elapsed review time to adjust future planning.
Common mistake
Treating review as invisible work. If the team only estimates what the AI generates, the sprint will fill up with unvalidated outputs. The R&I task prevents this false productivity by making the real bottleneck visible: human review, judgement, and integration.
Resources
🏦 Scrum en equipos con IA [Scrum in teams with AI, in Spanish]Skill Arena · Scrum Manager
🏦 SDD - Spec Driven Development en equipos ágiles [SDD - Spec Driven Development in agile teams, in Spanish]Skill Arena · Scrum Manager
📊 Guía didáctica SDD [SDD teaching guide, in Spanish]Resources · Scrum Manager
See also
Want to advance in agility? You can search for course and exam calls or go at your own pace by becoming a member of the Agile Club. This membership includes exclusive resources, e-learning classrooms, and access to Skill Arena: a space to practice and measure your agile skills at your own pace.