What is the typical flow from a database to a data lake, and what is the final step in the data pipeline?

The flow usually starts with a source database (OLTP) that feeds a data lake via CDC or batch extraction. Raw data lands in a bronze layer, then moves to a silver layer where it is cleansed and structured. In the gold layer, data is aggregated and optimized for analytics. The final step is exposing the gold layer to downstream consumers: either a data warehouse for SQL queries, a BI tool for dashboards, or an API layer for application consumption. In many architectures, the data lake acts as a central repository, and the data warehouse is built on top of it (lakehouse). The final step is often a semantic layer that abstracts the underlying schema, enabling self‑service analytics.