Databricks hit $6.9b in annualized recurring revenue, up 80% year over year. Snowflake sits at $5.3b ARR, up 34%. The gap between them was $490m in March. It is $1.6b now, and every quarter adds more distance.
AI is doing the pulling. Databricks' AI products run at $1.7b annualized, roughly 25% of total ARR, up from $1b just six months ago, growing faster than the company overall. The pattern repeats elsewhere: Salesforce paid $3.6b for Fin because the former Intercom's AI agent hit $100m ARR, also ~25% of its total, up 350%. The fastest-growing companies in software right now either sell AI directly, resell inference, or sit one step removed from the token path.
At a $134b private valuation, Databricks is larger than Salesforce and smaller than SAP, the only data-focused peer that outranks it. It outpaces every comparable company on growth: CrowdStrike at 26%, Shopify at 34%. The full piece is worth reading for Tunguz's framing of the 'first derivative of inference' thesis, which explains why proximity to the token path, not just AI branding, is what separates the accelerating companies from the rest.
[READ ORIGINAL →]