Team Role Assigner

Define responsibilities for your AutoML team

← Back to Hub

👥 Why Defined Roles?

Clear role definitions prevent duplication of effort, ensure accountability, and help teams work efficiently. Each role has specific responsibilities that contribute to the project's success.

🏃
Runner
Executes experiments

The Runner is responsible for executing training scripts, monitoring progress, and ensuring experiments complete successfully.

  • Execute training scripts
  • Monitor TensorBoard logs
  • Report progress to team
  • Troubleshoot runtime issues
  • Save and organize artifacts
🛠
Maintainer
Manages the repository

The Maintainer manages the Git repository, handles merges, ensures code quality, and maintains documentation.

  • Manage Git branches
  • Review and merge PRs
  • Maintain README and docs
  • Resolve merge conflicts
  • Tag releases
📊
Analyst
Interprets results

The Analyst interprets experiment results, creates visualizations, and provides insights to guide the team's decisions.

  • Analyze training metrics
  • Create result visualizations
  • Compare experiment runs
  • Identify trends and issues
  • Write results summary
🔎
Reviewer
Ensures quality

The Reviewer ensures code quality, validates results, and provides constructive feedback on all team deliverables.

  • Review code changes
  • Validate experiment results
  • Check documentation accuracy
  • Provide constructive feedback
  • Approve PRs

📝 Assign Team Members

✅ Role Checklists

  • Set up execution environment Install dependencies, configure hardware (MPS/CUDA)
  • Run baseline experiment Execute training with default parameters
  • Monitor training progress Check TensorBoard, watch for errors
  • Run HPO experiments Execute Optuna trials, track results
  • Save artifacts Export models, logs, and configs to designated location
  • Initialize repository Create repo, set up branch protection
  • Create README.md Document project setup and usage
  • Set up directory structure Create configs/, results/, docs/ folders
  • Review and merge PRs Ensure code quality before merging
  • Tag final release Create version tag for submission
  • Set up metrics tracking Define what to measure and how
  • Analyze baseline results Document initial performance
  • Compare HPO trials Identify best hyperparameters
  • Create visualizations Training curves, comparison plots
  • Write results summary Document findings and recommendations
  • Review eval protocol Ensure methodology is sound
  • Review code PRs Check for bugs, style, best practices
  • Validate results Verify metrics are correctly computed
  • Check documentation Ensure README and runbook are accurate
  • Final approval Sign off on submission readiness

📄 Team Roster

Copied to clipboard!