Getting Started with a Semantic Layer on Snowflake | AtScale

More than a decade into the big data era, data quality is still the top obstacle for data teams. Over 56% of data professionals cite poor data quality, incomplete, inconsistent, or outdated, as their most frequent challenge, and that instability now ripples into every dashboard, report, and AI model built on top of it, according to dbt Labs’ 2025 State of Analytics Engineering Report.

Despite major investments in powerful cloud data platforms like Snowflake, many businesses are still struggling to make the most of their data.

The problem is usually that it’s difficult to bridge the gap between the data in Snowflake and the tools that business users work with (such as Power BI, Tableau, or Excel). Many business users fall back on static data extracts (such as TDE files for Tableau), creating multiple data silos and fragmented insights.

The solution? Adding a semantic layer into the data stack between your Snowflake cloud and your business intelligence tools.

What is a semantic layer?

A semantic layer takes your business’s raw data and turns it into consistent, governed definitions, the metrics and dimensions that mean the same thing everywhere they’re used. It gives you a single source of truth: one definition of revenue, or churn, or margin, that stays the same whether a person is looking at it in Power BI, Excel, or Tableau, or an AI agent is querying it directly.

Instead of moving data in and out of the Snowflake Data Cloud, the semantic layer enables you to leave the data where it is. You can work within the semantic layer to prepare the data for analysis across any BI, ML, data science, or AI tool. Your data team no longer needs to remodel the data in your BI tools to make it usable; every user, and every AI agent, gets a live connection to the real data in Snowflake instead. As more data is added to Snowflake, changes are immediately propagated to everyone connected to the data models in the semantic layer, so your data is always accurate.

Why AtScale and Snowflake?

AtScale’s semantic layer connects directly to Snowflake through Snowflake Semantic Views, giving business users a governed, live path from Snowflake’s Data Cloud into the BI tools they already use. Here’s what that looks like in practice:

• A direct connection for Power BI and Excel. AtScale’s embedded XMLA endpoint (coming soon in private preview), built directly into Snowflake Semantic Views, connects Power BI and Excel straight to governed metrics in Snowflake. Under the hood, AtScale’s ACE engine powers that connection, so there’s no new middleware or separate connector to maintain. Read more in Inside the AtScale and Snowflake Partnership.
• Security and governance that scale with your data. AtScale’s semantic layer inherits Snowflake’s existing security model, and adds its own row-level and column-level security plus data masking, so governance grows with your data instead of needing separate infrastructure.
• Faster queries without added Snowflake cost. AtScale automatically creates data aggregates based on real query patterns, so reporting stays fast as usage grows, without scaling a separate analytics environment or your Snowflake warehouse spend.
• Room to grow beyond Power BI and Excel. AtScale supports DAX, MDX, Python, and SQL, so you can keep working in the BI tools your team already uses. When you need to extend that same governed semantic layer to Tableau, Databricks, or other tools, AtScale Enterprise adds that broader reach.

AtScale’s approach has been recognized outside the Snowflake partnership too. AtScale was named a Leader and Fast Mover in the 2025 GigaOm Radar for semantic layer platforms, following similar recognition in the 2024 GigaOm Sonar, with customers like Blue Yonder and Papa Johns already relying on AtScale in production.

See AtScale’s embedded XMLA endpoint for Snowflake Semantic Views in action, and explore how to bring governed, live semantic layer access to your Snowflake environment, at snowflake.atscale.com.