RetailX: A Self-Service Multi-Tenant Data Engineering Platform
We present RetailX, a turnkey data-engineering platform on commodity hardware: Apache Kafka ingests CSV, API, CDC, and visitor-pixel events; a Python consumer writes Parquet to the data lake; DuckDB serves as the warehouse; dbt transforms and tests the data; and Airflow schedules the pipeline every 10 minutes — all exposed through a 25-job terminal wizard and FastAPI REST API.




