Interview, Fireside Chat
Snowflake vs Databricks: The AI Data War | CEO of $SNOW
- Sridhar Ramaswamy contrasts Snowflake's trajectory with his experience at Google, where the Search Ads product grew from $1.6 billion in revenue in 2003 to $100 billion by 2018, noting that a $100 billion revenue plan drafted in 2007-2008 seemed "preposterous" at the time.
- He attributes such exponential growth to "35% odd growth compounded over and over again" over a decade or more, advocating for a long-term vision rather than short-term tactical fixes.
- Ramaswamy identifies the current AI supercycle as being in its very early stages, likening the industrialization of thought to a phenomenon spanning centuries rather than a discrete event.
- He estimates that over 75% of the stock market gains in 2025 (approx. $6 trillion) were driven by AI and AI-adjacent companies.
- He asserts that while value is currently accruing to NVIDIA, hyperscalers, and frontier model makers, it remains unclear how quickly open-source models will catch up and erode the moats of current leaders.
- Value creation is shifting from pure model and hardware development to practical application layers, with nearly a billion users now creating value via ChatGPT in unexpected ways.
- Key areas of immediate value creation include coding agents, enterprise customer support, and conversational data analysis (Snowflake Intelligence).
- He notes that old chips and models remain highly valuable for inference and simple tasks, even as new frontier models enable previously impossible capabilities like autonomous coding agents.
- Strategic positioning for Snowflake involves a dual focus on enterprise-grade reliability (disaster recovery, governance) and accelerating velocity to compete with fast-moving private sector players.
- Ramaswamy explicitly stated the company "should be asking ourselves why we are not growing a whole lot faster," acknowledging a competitive "wartime environment."
- He predicts that companies relying heavily on seat-based licensing and human-intensive service models will struggle against cheaper, more efficient AI-native alternatives.
- The acquisition of Observability platform Observe is designed to integrate telemetry data into Snowflake's AI agent ecosystem, enabling engineers to query system health and performance issues naturally.
- The deal brings a running business product into the fold, distinct from previous talent-focused acquisitions, allowing the Observe team to leverage Snowflake's scale and engineering resources.
- Ramaswamy plans to have the Observe team report directly to him to ensure they retain their "life force" and agility while integrating with the larger platform.
- Leadership traits for navigating the current volatile environment are defined by the combination of "drive" and "malleability," with a specific emphasis on the ability to change one's personal and professional behaviors.
- Ramaswamy notes that cultural change is best spread through "iconic people" rather than top-down mandates, citing the influence of co-founder Benoit Dageville's adoption of coding agents.
- He emphasizes that the organization has transformed from a long-planning database company to one operating iteratively, driven by the expectation that users and engineers can interact with data via natural language and APIs.
- Financial advice for upcoming trillion-dollar IPOs centers on managing market psychology and avoiding the "loss aversion" that can destabilize a company when valuations fluctuate.
- He advises CEOs to "nudge" valuations to maintain stability and recommends structuring IPOs to attract long-term institutional investors rather than volatile retail or hedge fund capital.
- Ramaswamy dismisses the existence of "secrets" regarding the company's strategy, stating, "you see is what you get," while acknowledging the intense personal discipline required, describing his lifestyle as "monk mode" with a focus on work, family, and gym, eschewing traditional hobbies.
- He highlights the high stakes for customers, noting that CIOs often bet their careers on Snowflake implementations, necessitating a stewardship approach to investor and customer relations.