Data Engineer — bridging classic data warehousing with the modern lakehouse.
I spent ~5 years deep in the Oracle DWH ecosystem (PL/SQL, ODI, dimensional modeling, SCD) building and optimizing enterprise data pipelines, primarily in the insurance domain. Now I'm carrying that foundation into the modern data engineering stack — CDC, streaming, and lakehouse architecture.
- Data Warehousing · Oracle, PL/SQL, ODI, star schema, SCD Type 2, MLOG-based CDC — production ETL/ELT at enterprise scale
- Modern Data Engineering · Kafka, Debezium (CDC), Spark Structured Streaming, dbt, Apache Iceberg, Airflow
- Lakehouse Architecture · End-to-end pipelines from source CDC to a queryable, ACID lakehouse with BI + ML on top
- Domain Expertise · Deep experience in insurance (claims, provisions, benefits, reconciliation)
🚀 Featured Project — ecommerce-realtime-pipeline
A self-hosted, real-time e-commerce lakehouse, built end to end:
Postgres → Debezium (CDC) → Kafka → Spark Structured Streaming → MinIO/Iceberg → dbt → Airflow → Superset
- Real-time CDC with soft-delete handling and LSN-based deduplication
- Medallion architecture (bronze/silver/gold) on Apache Iceberg — ACID, schema evolution, time travel
- ML layer — fraud detection, demand forecasting, customer segmentation, churn prediction (feature store in dbt, orchestrated by Airflow)
- AI access layer — an MCP server enabling natural-language analytics over the lakehouse
Oracle PL/SQL ODI Python SQL dbt Apache Spark Apache Kafka Debezium Apache Iceberg Apache Airflow Superset Qlik Sense PostgreSQL Docker MinIO
Most of my earlier work lived in enterprise Oracle/ODI repositories — my recent GitHub activity reflects my move into the open, modern data stack.
LinkedIn: https://linkedin.com/in/deniz-isik-ofc/
E-mail: denizsk977@gmail.com