COVER Tech explains Snowflake data platform and access-control practices
Summary
COVER Tech has published a guide to company-wide data use on Snowflake and access-control practices. It covers ETL, dbt modeling and per-project workspaces with RBAC roles for safe, self-serve analysis.
COVER Corporation has updated its technical blog COVER Tech with a post by engineer U from the CTO Office data infrastructure team.
The article outlines how COVER centralizes data from its own services such as the holoplus app and official e-commerce, plus platforms like YouTube, into Snowflake as a company-wide data platform. Raw data is modeled with dbt, with automated checks for quality and freshness using dbt test and Source Freshness.
The second half explains access-control practices for safe and open analysis. Each project or team receives its own database workspace, while the platform team controls access to base processed layers. Permissions are packaged as READ and READWRITE roles under Snowflake RBAC, with workspaces, roles and grants managed as code in Terraform. Because Snowflake ties features such as TASK, Stream and Streamlit apps to database objects, workspace-level control also governs feature use.
Looking ahead, the team wants to introduce per-workspace resource monitors so users can manage compute costs themselves while keeping analysis free and safe.
Full post in Japanese: https://coveredge.cover-corp.com/list/5273
Read the original