It’s been barely three months since Sridhar Ramaswamy, a former Google executive, took the CEO reins at Snowflake, and one major change under the new boss is already easy to spot: the decade-old company, known since long under the name “Data Cloud”. now presents itself as the “AI Data Cloud”.
The rebranding may seem like a simple marketing exercise in an era of widespread AI hype, but Ramaswamy is convinced that AI and data are inextricable parts of Snowflake’s business and future success. . AI now “permeates everything that happens in Snowflake,” Ramaswamy said. Fortune in an interview ahead of Snowflake’s annual summit in San Francisco this week. And AI has become “the nervous system” that connects companies internally, he said, helping them create and manage data, but also allowing companies to exchange information in real time through the cloud with other companies.
Ramaswamy rose to Snowflake’s top job in February after Frank Slootman unexpectedly announced his retirement. He had big shoes to fill: Slootman, a tech industry veteran, led Snowflake through an IPO that transformed it into a public company with a market cap of $75 billion. The company’s headcount more than tripled under Slootman as Snowflake expanded its client list and leaned into competition with fellow data startup Databricks.
Compared to Slootman, who liked to quote World War II General George Patton and whose leadership was described as “brash” and “pragmatic,” Ramaswamy comes across as low-key and diplomatic. But the challenges Ramaswamy faces are no less urgent. On the one hand, Wall Street is still mourning Slootman’s departure; Snowflake shares plunged on news of the management change, and the company, which has yet to turn a profit, now has a market capitalization of $46 billion.
During this time, the company faced a difficult news cycle. On Monday, Snowflake pushed back against claims by Live Nation, Ticketmaster’s parent company, that hackers had accessed data stored through Snowflake. According to Snowflake, the company found no evidence that the hack was the result of a vulnerability, misconfiguration or malicious activity within its product and, in a joint statement with other companies, said a targeted hacker campaign to exploit multi-factor authentication appears to be emerging. be the cause.
For Ramaswamy, the Snowflake Summit in San Francisco provides a platform to convince investors and rally customers as he argues that this is “the year of AI.”
One of Snowflake’s key announcements at this week’s summit is an expanded partnership with Nvidia that allows customers and partners to build custom AI data applications in Snowflake, powered by Nvidia AI Enterprise and Cortex AI software, the LLM and Snowflake’s vector search service.
In a virtual conversation with Ramaswamy on Monday evening, Nvidia CEO Jensen Huang said that companies need to “take all of the most important processes that they do, capture them in a data wheel and turn them into AI of the company to drive this steering wheel even further.” .”
Could Snowflake be the next Nvidia?
While some wonder whether Snowflake’s AI efforts could catapult the company into an Nvidia-like stratosphere of revenue and stock market value, Ramaswamy seems firmly rooted in the here and now.
“I once worked for a billion-dollar company, at Google,” he said. “It takes a lot of work, I think it took almost 15 to 18 years before Google became a billion-dollar company. It’s just a very long journey.
Snowflake is betting that Ramaswamy’s experience and ability to harness the power of data and AI will put the company on a path to greater success. After running Google’s $115 billion advertising business in the 2010s, Ramaswamy launched Neeva, an AI-powered search engine that Snowflake acquired a year ago. As part of the acquisition, Ramaswamy joined Snowflake as senior vice president of AI, intended to help integrate Neeva’s Large Language Model (LLM) technology with the large amounts of structured and unstructured data hosted on the Snowflake platform.
“It’s an area I know very well,” he said. “My doctorate was on databases. I’ve worked in data most of my life.
Neeva’s LLM expertise has proven to be of great value to Snowflake in helping clients glean structured insights from unstructured data – asking questions and getting answers reliable from commercial data. “A natural language interface for all the information we want to give or consume is a huge potential of AI,” he explained. “AI and Snowflake mean you can move from unstructured information to structured information; this seamless access to information offers enormous potential for a company like Snowflake. »
When it comes to AI, Ramaswamy says, Snowflake is focused on creating value. “Our job is not to sell promises,” he said. “Our goal is to get paid when our customers consume computers on Snowflake and pay us for it. »
Nonetheless, he said he believes Snowflake is “well-positioned” to benefit from future customer obligations using AI. “Our customers don’t buy GPUs, they don’t buy guaranteed contracts,” he said. “They create bonds, they use what they need from Snowflake and they make money when they get value from it: it’s a very consistent business model. And our model is largely based on what we are able to provide to our customers. I think it’s a good thing.
A huge opportunity in the data space
Snowflake has also been in the news recently over rumors of potential acquisitions, including a billion-dollar deal for AI model developer Reka AI that reportedly went off the rails. Snowflake has confirmed the acquisition of TruEra, which evaluates and monitors LLM applications and machine learning models in production.
“We tend to be quite strategic” about acquisitions that fill necessary niches, said Ramaswamy, who would not comment on the Reka rumors. “Neeva came into Snowflake because Frank and our owners realized that search was going to play a critical role for AI, and we had deep search expertise and AI expertise. At the same time, TruEra’s skills around so-called “observability” will be “important in the world we’re moving toward,” he said.
Ramaswamy is most interested in seizing the customer value opportunity that he believes lies before Snowflake. Cloud computing, he said, is estimated to be worth more than $2 trillion over the next decade. “A lot of that will be spent on data.”
And AI, he added, will be a key addition to extracting customer value from this cloud equation.
If Snowflake is the nervous system that provides connectivity between different cloud applications and massive amounts of data, AI may be the brain that can make those connections for the enterprise.
“There is a huge amount of value to be created for anyone who can solve this problem,” Ramaswamy said.