jyablonski / nba_elt_mlflow / 26416413718

91%

push

github

feat(ml): add dashboard refresh and post-run health checks (#33)

## Description

After ML predictions commit, the pipeline refreshes the dashboard
in-memory
snapshot and validates `/internal/health` before finishing. Failures
alert
Slack; high memory usage is warning-only.

### Added

- `src/dashboard_refresh.py` with refresh, health validation, and
retries
- Unit tests for dashboard checks and `run_pipeline()` orchestration
- Testcontainers Postgres for integration tests; `ty` in CI

### Updated

- `src/app.py` to run post-pipeline checks and send structured Slack
alerts
- CI/dev tooling (Python 3.14, sklearn 1.7 pin, pytest/coverage config)
- README and Makefile; removed docker-compose test postgres setup

### Deleted

- `docker/docker-compose-test.yml` and `docker-compose-postgres.yml`

164 of 182 new or added lines in 3 files covered. (90.11%)

271 of 293 relevant lines covered (92.49%)

0.92 hits per line