ML Workflow

Purpose

Design end-to-end ML workflows covering experiment tracking, feature engineering and storage, model training pipelines, model serving and deployment, A/B testing for models, and monitoring for data and model drift. Produces a workflow architecture, tool selection rationale, and operational runbook.

Scope Constraints

Reads ML code, configuration files, experiment logs, and infrastructure specs for analysis. Does not train models, execute experiments, or deploy to production.

Inputs

• ML problem type (classification, regression, ranking, recommendation, NLP, CV)
• Data sources and feature candidates
• Model complexity range (linear/tree-based vs deep learning)
• Serving requirements (batch predictions, real-time inference, edge deployment)
• Team size and ML maturity (first model vs established ML platform)
• Infrastructure constraints (cloud provider, GPU availability, budget)