Keep pulling the thread on Sridhar Ramaswamy.
In early 2023, Snowflake developed a credible Mixture-of-Experts (MoE) foundation model but pivoted away from this strategy after concluding it lacked the capital to compete with labs like OpenAI and Anthropic.
Sridhar Ramaswamy predicts the line between agentic systems and traditional software applications will become "bloody" as agent capabilities expand.
Snowflake Intelligence is an opinionated agentic platform specifically focused on extracting value from data within Snowflake, rather than being a general-purpose agent platform like those from cloud service providers.
Sridhar Ramaswamy's thesis for joining Snowflake was that a data platform spanning multiple cloud service providers has the potential to become as large as a CSP itself.
Snowflake's internal sales data assistant, codenamed Raven, was a precursor to the Snowflake Intelligence platform and is used to consolidate sales dashboards and customer information into a single interface.
Sridhar Ramaswamy believes no software company can feel secure in its market position due to the rapid pace of change driven by AI.
Snowflake has deployed coding agents to its solution engineers, which has dramatically lowered the time required to create customized, synthetic data-based demos for specific customers.
Snowflake has been working on a partnership with Microsoft for the past two years to improve product integration, including with Microsoft Fabric, and to enhance go-to-market collaboration.
Sridhar Ramaswamy predicts the advertising business model will persist and reinvent itself for chat-based interfaces.
Snowflake Intelligence is priced on a consumption basis, allowing customers to pay for what they use rather than through a subscription model.
Approximately half of the qualifying Fortune 2000 companies are Snowflake customers.
Prior to Sridhar Ramaswamy becoming CEO, Snowflake's organizational structure had seven to ten layers of teams between an engineer building a feature and the customer using it.