mle-workflow - Production machine-learning engineering workflow
Guides production ML systems through data contracts, reproducible training, evaluation, deployment, monitoring, and rollback.
Tags
Updated: 2026-10-15Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Define data contracts
- Design reproducible training
- Evaluate model quality
- Package inference artifacts
- Design serving paths
- Plan model rollouts
- Monitor production models
- Specify rollback paths
- Analyze model errors
- Review ML risks
Inputs
- ML system requirements
- Training data
- Feature definitions
- Label definitions
- Evaluation metrics
- Serving constraints
- Deployment configuration
- Monitoring requirements
Outputs
- Data and metric contracts
- Training pipelines
- Evaluation reports
- Versioned model artifacts
- Inference APIs or batch jobs
- Rollout plans
- Monitoring dashboards
- Rollback plans
- Error analysis reports
- Refresh plans
Requirements
- Machine-learning project context
- Access to relevant code and data
- Applicable testing and deployment tooling