Who checks the work?
Give reviewers the context and time to judge the result. Agree which checks are needed before the work is used or shared.
Role guides
Use these guides to discuss what people do, check and own as your team works with AI agents.
HELM 1.0.1 · Updated
When an agent takes on part of a task, review the responsibilities around it. Who provides the context? Who checks the result? Who decides whether it is ready to use?
The roles below are starting points for that conversation. Adapt the responsibilities to your team’s work, workload and risk. Related responsibilities may sit with the same person.
The six principles explain HELM’s approach. Use the Practitioner Guide for day-to-day work and the Leadership Guidefor team decisions and learning.
Choose a recent task and discuss the checks, expertise and results it needed.
Give reviewers the context and time to judge the result. Agree which checks are needed before the work is used or shared.
A task may cross several parts of the product. Identify who can judge each part and involve them when a change needs their skills.
Look at the effect on users alongside delivery volume, quality and effort. More output alone does not show that work improved.
Responsibilities
| Role guide | Related HELM responsibilities | Focus | |
|---|---|---|---|
| Software EngineerBackend Engineer, Full-Stack Engineer | Product Engineer | Build, review and maintain software with AI agents | |
| Staff / Principal EngineerPrincipal Engineer, Solutions Architect | AI Architect | Set technical direction and clear boundaries for agent-assisted work | |
| Engineering ManagerDevelopment Manager, Team Lead | Engineering Manager | Support team learning and review the results of agent-assisted work | |
| SRE / DevOps EngineerDevOps Engineer, Infrastructure Engineer | AI Reliability Engineer | Monitor agent behavior, cost and recovery alongside service reliability | |
| QA Engineer / SDETSDET, QA Lead, Test Engineer | QA Engineer + Evaluation Lead | Check agent output and the product people use | |
| Platform / Infrastructure EngineerInfrastructure Engineer, DevOps Engineer | Platform Engineer | Provide dependable shared tools for agent-assisted work | |
| Product ManagerProduct Owner, Business Analyst | Product Manager | Make goals clear and help your team build what users need | |
| Product DesignerUI Designer, Product Designer, Interaction Designer | Product Designer | Design clear experiences and review agent-generated interfaces |
Shared skills
Use the same five competencies across roles. The examples and practice activities change with the work.
For simple explanations and everyday examples, explore shared skills for working with AI.
Choose a useful outcome and the simplest approach. Decide what to delegate and when a person needs to step in.
Give people and agents the goal, context, constraints and a clear idea of a good result.
Check AI-assisted work against facts and requirements. Make uncertainty visible.
Take responsibility for the work you use or share. Respect decision boundaries and follow through on problems.
Use feedback to improve, practice the skills you need to check agent work, and share useful lessons.
Role guides
Related responsibilities: Product Engineer in HELM
Build, review and maintain software with AI agents
Related responsibilities: AI Architect in HELM
Set technical direction and clear boundaries for agent-assisted work
Role guide in HELM
Support team learning and review the results of agent-assisted work
Related responsibilities: AI Reliability Engineer in HELM
Monitor agent behavior, cost and recovery alongside service reliability
Related responsibilities: Platform Engineer in HELM
Provide dependable shared tools for agent-assisted work
Related responsibilities: QA Engineer + Evaluation Lead in HELM
Check agent output and the product people use
Role guide in HELM
Make goals clear and help your team build what users need
Role guide in HELM
Design clear experiences and review agent-generated interfaces