mle-workflow
Production machine-learning engineering workflow for data contracts, reproducible training, model evaluation, deployment, monitoring, and rollback. Use when building, reviewing, or hardening ML systems beyond one-off notebooks.
- installs 8w
- 1,562
- 30-day movement
- starts with the next reading
- Related entries
- 4
- Connections
- 0
An agent skill that encodes a production machine-learning engineering workflow: prediction and data contracts, reproducible training pipelines, evaluation gates, packaging for serving, monitoring, and rollback. It is structured as a SKILL.md with sections for when to activate, scope calibration, a six-step core workflow, review checklists, and supporting practices like error analysis and decision loops.
Reach for it when model work needs to move beyond one-off notebooks into a reviewable, production-grade ML system with explicit contracts and quality gates.
Use it to
- Plan or review a production ML feature or pipeline
- Convert notebook experiments into reproducible training code
- Define data and prediction contracts before model code
- Set promotion criteria and evaluation gates before training
- Run an error-analysis and monitoring loop after changes
For Engineers building, reviewing, or hardening production ML systems
- Host repository
- affaan-m/ECC
- Installs, lifetime
- 2,900
- Installs, 8 weeks
- 1,562
- Licence
- MIT
- Host stars
- 261k
- Host language
- JavaScript