Some customers are spending hundreds of millions of dollars on AI.
Databricks releases products to compete with ClickHouse, Splunk & Oracle
Databricks is seeing higher growth as AI agents assist with data analysis, but all that activity is significantly increasing costs.
Inference surpassed training spend in 2026; production deployments hit 500–1,000% cost overruns, especially with agentic workflows. Efficiency—not model size—is the differentiator; cost visibility per feature and routing routine tasks to cheaper models become critical.
Databricks surpasses $5.4B ARR with over 80% sales growth, but AI infrastructure costs push gross margins from above 80% to the mid-70% range ahead of a
After months of training hype, our AI feature went live and the cloud bill tripled; per‑token pricing blew past a $200 k budget, leaving finance scrambling.