Initialization of a new data platform

  • Practice
  • Team

1 data architect & 2 data engineers

  • Technical environment
A data platform
that is both observable and accessible
A data warehouse
that is rich and well-documented
A data model
that is strong and future-ready
  • After migrating the the Google cloud (GCP), need to rationalize their data platform
  • Decommissioning of both « legacy» data warehouses (on Oracle and Snowflake), replaced by a new one built on BigQuery
  • Strong will of initializing an operating model and a change management plan allowing a wider audience of data citizens to understand and leverage the data
  • For the Phase 1, focused on the data foundation, reduce the impact on the reporting tools (Tableau and SAP BI4) as much as possible
  • Build a solid foundation, allowing the group to move toward its transformation ambitions
  • Starting with a scoping phase, where we identified the scope, defined the high-level target model and documented the development best practices for dbt and BigQuery
  • Initialization of a data catalog, to make the full scope accessible to many
  • Development of a range of dbt models to build the bronze, silver and gold layers of the new datawarehouse
  • Development of a python toolbox to generate the bronze models and some of the silver ones from the data catalog
  • Set up elementary, to increase the observability of the platform
  • A data platform that is both observable and accessible to a wide audience
  • A rich and well-documented data warehouse
  • A strong and future-ready data model
  • An easy migration of the reporting, with only minor adjustments