Skip to content

Implementation Status

This is the canonical source of truth for implementation readiness. All claims in other pages must be consistent with this page.

Last updated: April 28, 2026 (v1 cycle Phase 0–2 + Phase 4)

v1.0 scope: see Requirements — Definition of Done and Roadmap — v1.0 Demo Track for the binding 1–2-week scope. Items below not part of v1.0 stay at their current status until post-v1.

Legend

  • ✅ Operational — Implemented, tested, and working in practice
  • 🚧 Partial — Partially implemented or requires manual steps
  • 📋 Planned — Designed but not yet implemented

Implementation Matrix

Component Status Notes
Data Engineering
Airflow ETL ✅ Operational Scraping + PostgreSQL ingestion
MinIO Storage ✅ Operational S3-compatible object storage
DVC Versioning ✅ Operational Data + model artifacts tracked
PostgreSQL ✅ Operational Canonical data store
ML Pipeline
Feature Engineering ✅ Operational stats_matches.py — time-windowed stats
DVC Pipeline ✅ Operational dvc.yaml orchestration working
MLflow Tracking ✅ Operational Experiment logging functional
Train/Test Splitting ✅ Operational Time-based + CV folds
Model Training 🚧 Partial Baseline + XGBoost classifiers wired; smoke parameters active by default (classification.fracs_for_train=[0.001, 0.002], tuning.n_trials=2); production-scale run is a v1.0 deliverable
Model Registry 🚧 Partial Registration automated; Staging→Production gate is manual
Serving
FastAPI App ✅ Operational Routers, middleware, lifespan; CORS allow-list driven by CORS_ALLOWED_ORIGINS (default empty = no cross-origin)
POST /predict ✅ Operational Sync inference via Celery ml queue, 30 s timeout
GET /predict/{match_id} ✅ Operational Lookup from batch_inference parquet output
POST /predict/async/ ✅ Operational Async Celery job, returns task_id for polling
GET /predict/model/info ✅ Operational MLflow model metadata from registry
Request Validation ✅ Operational Pydantic schemas in src/app/schemas/predict.py
Model Loading ✅ Operational Lazy-loaded once per worker process via PredictionService
Batch Predictions API 📋 Planned DVC batch_inference stage exists; no HTTP batch endpoint
Streamlit UI ✅ Operational src/ui/app/main.py (livescores), pages/1_Predictions.py (1×2 outcome), pages/2_Model_Metrics.py (champion/challenger metadata + metrics, read-only). Demo disclaimer rendered on every page.
Monitoring
Prometheus Metrics ✅ Operational GET /metrics, _PrometheusMiddleware, 8 counters/histograms/gauges
Service Health ✅ Operational GET /healthcheck/, liveness probes
Celery Queue Stats ✅ Operational GET /monitoring/celery/queues, /celery/workers
Task Status Polling ✅ Operational GET /monitoring/task_status/{task_id}
Grafana Dashboards 📋 Planned Prometheus exporting; dashboards not yet deployed
Evidently Drift Detection 📋 Planned Not integrated
Alerting Rules 📋 Planned Runbooks documented; rules not deployed
Infrastructure
Docker Images ✅ Operational Multi-stage builds for API + workers
K8s Manifests ✅ Operational Deployments + Services + ConfigMaps
Helm Charts ✅ Operational Values + templates parameterized; nginx-ingress rate-limit (rps/burst/connections) configurable via ingress.rateLimit
GitLab CI ✅ Operational Build + test + deploy pipeline
Secrets (SOPS) ✅ Operational age encryption for sensitive data
Quality
pytest Framework ✅ Operational 316 tests collected (unit, property, service, contract, load); see docs/planning/20260428_test_v2.md for coverage gap matrix
Unit Tests ✅ Operational tests/unit/ — splitting, schemas, preprocess
Property Tests ✅ Operational tests/property/ — Hypothesis: features, splitting, metrics
Service Tests ✅ Operational tests/service/ — prediction service, Celery tasks
Contract Tests ✅ Operational tests/contract/test_pipeline_contracts.py
Load Tests ✅ Operational tests/load/locustfile.py — Locust load scenarios
Integration Tests 🚧 Partial API mock tests; no live MLflow/Celery required in CI
Pre-commit Hooks ✅ Operational ruff + basic linting
Data Validation ✅ Operational Great Expectations suites at raw / finished / future / features stages

Known Limitations

  1. Model promotion — Registration is automated; Staging→Production gate requires manual approval. No automated promotion policy exists.
  2. Batch HTTP endpoint — batch_inference features are computed by DVC pipeline only; no HTTP endpoint for batch requests.
  3. Grafana — Prometheus is exporting 8 metrics. Dashboards are not yet deployed.
  4. Evidently — Drift detection is architecturally designed; not yet integrated.
  5. Alerting — Alerting rules are documented in runbooks; not deployed in Alertmanager.
  6. Integration tests — All CI tests use mocks. No live Celery/MLflow dependency in CI.
  7. Feature store — Features are file-based Parquet; no dedicated online store.
  8. API authentication — All public endpoints are unauthenticated. Access is TLS-only.

See Architecture Limitations for deployment-level constraints.

How to verify

# Reproduce the ML pipeline
dvc pull && dvc repro

# Inspect experiments
mlflow ui --port 5001

# Run the test suite
pytest tests/ -q

# Check API health
curl http://localhost:8000/healthcheck/

# Check Prometheus metrics
curl http://localhost:8000/metrics

See Quickstart for the full reproducibility path.