Databricks sales growth tops 80%, but margin are shrinking from swarm of AI agents

Databricks is experiencing a remarkable surge in revenue, capturing the attention of businesses eager to utilize its data analytics tools amid the ongoing AI revolution. However, the increasing reliance on AI agents for data management is imposing higher operational costs, which is impacting the company's profit margins. In a recent interview at the Data and AI Summit in San Francisco, CEO Ali Ghodsi explained, "The consumption-based business model is evolving with the rise of agentic AI. The influx of queries generated by these agents is significant. Our agent platform not only drives revenue but also amplifies overall consumption." Databricks reported an impressive annualized revenue growth of over 80% year-over-year, reaching $6.9 billion, a substantial increase from the previous $5.4 billion in the last fiscal quarter. With a private market valuation soaring to $134 billion, Databricks now surpasses its competitor Snowflake, which has a market capitalization of approximately $83 billion and annualized revenue of $5.6 billion. Despite its high valuation, Databricks remains privately held while other tech companies pursue initial public offerings. Notably, SpaceX made headlines last week with the largest debut on record, achieving a $2 trillion market cap on its first trading day. Meanwhile, AI developers OpenAI and Anthropic have also filed for confidential IPOs. While often associated with AI model developers, Databricks occupies a distinct position in the market. Its innovative tools, such as Genie for business data inquiries and Agent Bricks for custom AI application development, are gaining traction. However, the company is also facing the need to invest significantly in the underlying AI models to support these products. Ghodsi did not disclose the current gross margin but acknowledged it is expected to decrease. Notably, Databricks' annual revenue from AI-related products rose to $1.7 billion, up from $1.4 billion earlier this year. A notable trend among companies is the shift from excessive token usage to a more measured approach, termed "value-maxxing," which aims to optimize efficiency and control costs while leveraging AI capabilities. Large enterprises are eager to access advanced AI models, such as Anthropic's Mythos, but they also seek cost-effective solutions for routine tasks, often turning to simpler open-source models. Additionally, Databricks is responding to customer demands for diverse AI options, including the growing popularity of Chinese models. In pursuit of expansion, Databricks is refining its offerings for specific industries. The company recently announced its entry into the cybersecurity sector with the launch of Lakewatch software and revealed plans to acquire Panther, a security startup valued at $1.4 billion in 2021, alongside introducing CustomerLake software for marketing data management.