Skip to content

ML System Design Coverage

Status: ✅ Current as of May 2026 — coverage map complete. See Implementation Status for per-component readiness.

This page is a reading guide for technical interviewers and ML engineers who want to evaluate the project through the lens of ML System Design. It maps each major design topic to the specific files, modules, and decisions where it is addressed in this project.

The project covers all 15 topics. 11 are fully implemented; 3 are partially implemented (infrastructure gaps on the roadmap, not design gaps).


Coverage Map

# Topic Coverage Key Pointers
1 Problem & Solution Space problem.md, requirements.md
2 Preliminary Research baseline.md, limitations.md
3 Design Document requirements.md, tradeoffs.md, adr/
4 Loss Functions & Metrics baseline.md, tuning.md
5 Datasets & Data Work data/index.md, data/contracts.md
6 Validation Schemes validation.md
7 Baselines baseline.md
8 Error Analysis 🚧 src/pipelines/error_analysis.py, limitations.md
9 Training Pipelines training-pipeline.md, dvc.yaml
10 Feature Engineering features.md, src/features/
11 Reporting 🚧 MLflow UI, reports/qmd/, Results
12 Integration model-contract.md, serving/index.md
13 Monitoring & Reliability 🚧 monitoring/index.md, src/pipelines/monitor_drift.py
14 Inference Serving & Optimisation inference-modes.md, performance.md
15 Model Management & Governance model-registry.md, model-registry.md#promotion-policy

Open gaps (🚧 topics)

Topic What is missing Roadmap item
8 — Error Analysis No formal slice-analysis by league / season / team; error_analysis.py computes ROI only Post-v1
11 — Reporting Dynamic ROI panel in Streamlit UI shows filtered historical ROI; dedicated model-quality metrics page (log-loss, calibration, champion vs challenger) not yet implemented Post-v1
13 — Monitoring Grafana dashboards and Alertmanager rules not yet deployed Near-term roadmap

See Roadmap and Status for current priorities.