A health-data lakehouse that turns messy FHIR, claims, wearable, and notes data into three trusted outputs — analytics marts, an ML feature store, and a vector index — with de-identification, OMOP standardization, and cross-engine parity testing that caught a real de-id gap.
A production-grade ELT pipeline that automates daily identification of high-value customers using Apache Airflow, dbt-spark, and Apache Iceberg.
A practical guide to building and orchestrating robust data pipelines using Apache Airflow, covering local, cloud (GCP), and Kubernetes deployments.