Databricks reports annual revenue up 60%+ to $1B in its fiscal year to January 2023 and plans to acquire Rubicon, an AI storage system startup in stealth mode
Context & Ripple Effects
This report marks the moment Databricks crossed its first billion-dollar revenue year — up more than 60% for the fiscal year to January 2023 — two years after the $1B raise at a $28B valuation. The Rubicon deal is a tuck-in into AI storage while the startup is still in stealth.
Read against the later arc, the move looks foundational: annualized revenue went from this $1B base to $3.7B by mid-2025, then through the $134B-valued raise with $5.4B annualized, and by mid-2026 $6.9B annualized with margins under pressure from AI agent usage. Owning the storage layer matters precisely because AI workloads are where those costs concentrate.
First-order effects
- Acquiring Rubicon brings an AI-native storage system in-house at stealth stage, extending the platform below analytics into infrastructure just as revenue crosses $1B.
- Crossing $1B in annual revenue validates the lakehouse bet made when the company was valued at $28B, strengthening its position against cloud-native rivals for enterprise data-and-AI budgets.
Second-order effects
- Controlling storage economics gives Databricks leverage on the cost side that later shows up as decisive: CEO Ali Ghodsi's admission that AI agent usage raises costs and lowers margins makes owned infrastructure a margin lever, not just a feature.
- Competing platforms face pressure to match full-stack depth across data, AI tooling, and now storage, pushing consolidation of point-solution startups into larger platforms.
Third-order effects
- If the pattern holds, data platforms absorb infrastructure layers through tuck-in M&A rather than building them, concentrating the AI data stack in fewer vertically integrated players.
- The funding structure behind this expansion — multibillion-dollar equity rounds layered with debt at ever-higher valuations — ties platform consolidation to sustained access to cheap capital, making the model sensitive to financing conditions.
The trend: Data-and-AI platforms are buying their way down the stack into storage and infrastructure, converting hypergrowth revenue into vertical control of the AI workload economics.