Databricks' pre-IPO round signals confidence that the AI-data flywheel is finally working

Databricks' pre-IPO round, which values the data-and-AI platform at a figure in the low-to-mid tens of billions of dollars, is the latest evidence that the company founded by the creators of Apache Spark and Delta Lake has successfully navigated the transition from a data-engineering tools vendor to an AI platform. The round comes at a moment when the company's revenue mix is shifting toward AI-specific products - model training, fine-tuning, RAG pipelines and inference serving - at a rate that analysts say is changing the underlying multiple at which Databricks should be evaluated. The company's product portfolio sits at the intersection of two of enterprise technology's most active investment areas: data management and AI development. Its Unity Catalog governs data assets across cloud environments; its Mosaic AI platform handles the full model lifecycle from training to production monitoring; and its Lakehouse architecture, which combines the cost advantages of data lakes with the query performance of data warehouses, has been adopted by thousands of organisations as a foundational data platform. This breadth means Databricks competes with Snowflake on analytics, with cloud providers on AI training infrastructure, and with specialised MLOps vendors - simultaneously. The competitive dynamics are complex but increasingly resolved in Databricks' favour at the enterprise level. The company has been signing larger and longer contracts, with multi-year commitments from Fortune 500 companies building proprietary AI systems on its platform. The Dolly open-source model release and subsequent DBRX model have bolstered its credibility in AI research circles and attracted enterprise buyers who want a vendor that understands model development from first principles. For Indian IT companies and the domestic data-platform market, Databricks is a significant reference point. Infosys, TCS and Wipro have all built service practices around the platform, and several large Indian conglomerates - in financial services, telecommunications and manufacturing - have adopted it for their internal AI buildouts. The India revenue contribution, while not broken out separately, is growing rapidly given the scale of data-engineering work being done by Indian engineering teams globally. What to watch: the IPO filing timeline and whether the company goes public in the second half of 2025 as widely anticipated, how Snowflake's competitive response affects enterprise contract dynamics, and whether the DBRX model release translates into material uplift in AI-platform subscription revenue. The competitive pricing dynamics between Databricks and Snowflake in enterprise renewal conversations will be one of the clearest real-time indicators of which platform is winning the data-and-AI workload battle.

Original source: Bloomberg