A unified 'Data Intelligence Platform'. Following its MosaicML acquisition, Databricks dominates enterprise AI by seamlessly combining data governance (Unity Ca
Databricks Mosaic AI is the data-intelligence approach to enterprise AI: lakehouse data, governance (Unity Catalog), model training/serving, vector search, and agent tooling in one platform, betting that AI quality is downstream of data quality, governed end-to-end.
Pricing: Consumption-based DBU pricing across compute tiers and AI services; enterprise contracts negotiated at scale (published pricing, mid-2026).
Databricks vs Snowflake (2026): Both converged on data+AI platforms from opposite shores. Databricks leans engineering-native: lakehouse, notebooks, training, Mosaic AI agents. Snowflake leans analyst-native: SQL-first Cortex AI, governed simplicity. Existing estate and team skill profile decide more than feature checklists.
Both converged on data+AI: Databricks leans engineering/ML-native (notebooks, training, lakehouse), Snowflake leans SQL/analyst-native with Cortex services. Existing estate and team skills usually decide.
Governance travels with the data: permissions, lineage, and quality controls apply to RAG and agents automatically, solving the 'AI exposes every data gap' problem at the platform layer.