Databricks and Snowflake Platform Convergence

This rivalry now matters less as a warehouse fight and more as a battle to own the full daily workflow of a data team. Snowflake started with the cleaner SQL warehouse that analysts could query quickly, while Databricks started with Spark jobs, notebooks, and ML pipelines for engineers. Each is now moving into the other side’s budget, which turns expansion products into the real scorecard, especially Databricks SQL, Mosaic AI, Snowpark, and Cortex.

• Databricks has already built meaningful revenue outside its original core. By late 2025, about 40% of Databricks revenue came from newer layers, with Databricks SQL and Mosaic AI each at a $1B run rate. That shows warehousing and AI are no longer side bets, they are central to how Databricks grows against Snowflake.
  1 sacra 2 sacra
• Snowflake is still much more dependent on warehouse spend. In the same period, about 94% of Snowflake revenue still came from warehousing, while Snowpark, Cortex, and Streamlit were much smaller. That makes Snowflake’s push into engineering and AI strategically necessary, because it needs more reasons for customers to keep workloads on platform.
  2 sacra
• The overlap is visible in adjacent tools that sit on top of both platforms. Governance vendors like Immuta build deep integrations with Snowflake and Databricks because large enterprises often run both at once. That is a sign that many buyers are no longer choosing one stack for everything, they are comparing feature by feature across the same accounts.
  4 sacra 5 sacra

The next phase is platform consolidation around AI and application building. Databricks is pushing upward from infrastructure into apps, agents, and databases, while Snowflake is pushing outward from analytics into engineering and AI services. The winner will be the platform that becomes the default place where enterprise teams store data, run models, build internal tools, and govern access in one loop.