Initial release (v1.0.0) of the sanitized ML-Based Risk Scoring pipeline on GCP.
Includes:
- Full Batch/ETL ingestion design
- Feature engineering patterns
- BigQuery ML model training + scoring flows
- Governance (IAM, CMEK, data contracts)
- Lineage + auditability
- Runbooks for replay, backfill, rollback
- JSON data contracts for transactions + risk scores
- Diagrams (Architecture + Dataflow)
- Full docs-only repo structure with ADRs
No client code or sensitive data. Fully safe for portfolio and recruiters.