Crunchbase and Snowflake Ventures Examine What Actually Scales in Enterprise AI

A divide is forming in plain sight, and it has little to do with who can access models. One side is still performing intelligence, stacking demos, polishing outputs, selling the appearance of capability. The other is dealing with the harder question of durability, whether any of it holds once it collides with infrastructure, governance, enterprise security, and scale. That separation is no longer theoretical, and the startup ecosystem is beginning to sort itself accordingly.

That tension sits underneath “What It Takes To Scale AI,” a virtual event hosted by Crunchbase and Snowflake Ventures on May 13. This is not a broad conversation about possibilities. It is a direct examination of where enterprise AI is gaining real traction and what the data says it actually takes to scale. At a time when capital is tightening around fewer, larger bets, the conversation moves away from surface-level fascination and into the operational reality shaping the next cycle of enterprise adoption.

Crunchbase Research Lead Gené Teare and Harsha Kapre, Head of Snowflake Ventures, bring two different lenses into the same pressure point. Gené Teare has spent years tracking global funding trends, unicorn creation, and the movement of investor conviction across emerging sectors. Harsha Kapre brings more than 20 years of experience across database and data management technologies, with direct visibility into the companies building on top of the Snowflake platform and the infrastructure decisions separating scalable businesses from temporary momentum.

Inside the session, Crunchbase and Snowflake Ventures pull from new findings tied to analysis of the Snowflake partner ecosystem. The numbers tell their own story. Since 2020, $113B has flowed into that ecosystem, with investor appetite increasingly concentrating around companies building secure, governed, enterprise-ready systems capable of moving beyond the pilot stage. The market is becoming more selective, and the startup ecosystem is starting to reward companies built on durable data foundations instead of narrative alone.

The conversation sharpens around 3 critical areas. First, where investor conviction is rising and which categories are pulling larger concentrations of capital. Second, which layers of the AI data stack are proving most critical as enterprises move from experimentation into scaled deployment. Third, what the next phase of enterprise AI actually looks like, including Snowflake Ventures’ perspective on agentic AI, enterprise readiness, and the signals separating scalable companies from the rest of the market.

The audience reflects the stakes. Venture capital investors. Corporate venture teams. Startup founders. Executive operators. Product, strategy, and innovation leaders trying to understand where enterprise demand is solidifying and where noise is still masking weak foundations. The value here is not access to another headline cycle. It is clarity. Shared language. A cleaner read on where infrastructure, capital, and enterprise adoption are beginning to align across the startup ecosystem.

Snowflake has consistently framed enterprise AI as a data problem before it becomes a product story. Crunchbase adds the market intelligence layer, tracking where money, momentum, and investor confidence are converging. Put those views together and the conversation becomes less about hype and more about filtration. Which companies can scale. Which architectures can hold. Which signals still matter once markets stop rewarding ambition without operational proof. momentum.