A semantic layer is a translation layer between raw data and users (both human and machine), provides a unified, meaningful, and user-friendly view of complex and often distributed raw data, enabling consistent understanding, easier access, and better governance of data across tools and teams. The semantic layer - Defines business-friendly terms and concepts (e.g., "Customer", "Revenue", "Order Date") rather than raw column names like `cust_id` or `txn_dt`. - Maps these concepts to the underlying data structures across tables, files, or systems in your data. The semantic layer often includes [[metadata]], relationships, rules, and [[ontology|ontologies]] to add meaning and context. Common semantic layers include - [Snowflake Cortex Analyst](https://www.snowflake.com/en/developers/guides/getting-started-with-cortex-analyst/): write a semantic layer in YAML to allow natural language queries over a Snowflake data warehouse - [Databricks AI/BI](https://www.databricks.com/product/business-intelligence) (Genie): A natural language interface that sits on top of [Databricks Unity Catalog](https://www.databricks.com/product/unity-catalog). AI/BI makes it possible for business users, even those who don't speak SQL or write code, to get trusted insights quickly and accurately. With Genie, users can self-serve and obtain answers to questions not addressed in their dashboards without having to learn how to use BI tools or rely on expert practitioners to create insights for them. - dbt Copilot (powered by MetricFlow) - Microsoft Fabric Copilot: Fabric supports Lakehouses and focuses on connecting data sources to Power BI dashboards Tableau Pulse ### Comparison: Cortex Analyst vs. Fabric Copilot | Feature | Snowflake Cortex Analyst | Microsoft Fabric Copilot | | :--- | :--- | :--- | | Philosophy | "Semantic First" | "Ecosystem First" | | The "Brain" | You define a strict YAML Semantic Model. It forces the AI to use your definitions. | It relies heavily on Power BI semantic models. If you have a Power BI dataset, Copilot uses that. | | Accuracy | Higher precision for structured data because of the YAML enforcement. | Good, but can be "chatty." Better at generating DAX (Power BI code) or Python code. | | Integration | Best for SQL-heavy teams. | Best for Teams/Office users. You can ask questions directly inside Microsoft Teams. | | Governance | RBAC on the tables + Semantic View. | Relies on Microsoft Entra ID (Azure AD) and Power BI RLS (Row Level Security). |