Snowflake built its reputation as the go-to cloud data warehouse for enterprises. Store your data, query it fast, scale on demand. That was the pitch, and it worked — the company hit $5.4 billion in revenue and reached a $134 billion valuation on the strength of that model.
But CEO Sridhar Ramaswamy says that chapter is closing. In a new interview on TechCrunch's Equity podcast, Ramaswamy made it clear that Snowflake is no longer content being the place where data lives. It wants to be the platform where data works — autonomously, through AI agents that plan, execute, and deliver results without waiting for a human to write SQL queries or build dashboards.
The transformation is already underway, and it signals a much larger shift in how enterprise software will function in the age of AI.
The Chatbot Era Is Over
Ramaswamy's central argument is blunt: the chatbot phase of AI was a stepping stone, not the destination. Chatbots answer questions. They summarise documents. They help draft emails. But they do not complete real work.
The next phase — what Ramaswamy and the broader industry call the "agentic era" — is about AI systems that can take multi-step actions on their own. Instead of asking a chatbot to explain your sales data, you ask an agent to build a board-ready presentation with the latest numbers, identify churn risks, and recommend next steps. The agent pulls the data, runs the analysis, formats the output, and delivers it — all without a data analyst in the loop.
This is not a theoretical vision. Snowflake launched Project SnowWork in March 2026 to make this a reality. The platform lets business users describe what they need in plain language, and AI agents autonomously execute the workflow end to end, drawing from the company's governed enterprise data.
From Storage Layer to Execution Layer
The strategic logic behind Snowflake's transformation is straightforward. For years, enterprises have poured data into Snowflake's platform — financial records, customer data, supply chain metrics, operational logs. That data is governed, secured, and structured. It is the single source of truth for thousands of organisations.
Snowflake's bet is that the company sitting on all that governed data is uniquely positioned to build the execution layer on top of it. AI agents need trusted, clean, policy-compliant data to function reliably. Without it, autonomous systems hallucinate, make errors, or violate compliance rules. Snowflake already has the data foundation — now it is building the intelligence and action layers above it.
This is what Ramaswamy means by "shipping with your data." The data warehouse is no longer the product. It is the foundation on which AI agents operate.
Hundreds of AI Features, Plus Internal Restructuring
Snowflake is not making this transition slowly. The company has shipped hundreds of AI features over the past year and restructured its internal teams to align with the new strategy. Key moves include the acquisition of TensorStax in early 2026 to accelerate agentic AI for data engineering, the launch of Cortex Code as an AI coding agent for developers, and the rollout of Snowflake Intelligence for natural language data queries.
Project SnowWork ties all of these pieces together by adding a workflow execution layer for non-technical business users. It comes with pre-built profiles for roles in finance, sales, marketing, and operations — each tuned to the workflows, metrics, and terminology specific to that function. The platform inherits Snowflake's existing security controls, including role-based access, data masking, and audit logging.
As of February 2026, Snowflake reported that 9,100 accounts were actively using its AI products — a number the company will use as a baseline to track how quickly the agentic platform gains adoption.
The Competitive Landscape Is Heating Up
Snowflake is not alone in this race. Microsoft, Salesforce, ServiceNow, and Nvidia have all launched their own agent-based platforms aimed at turning enterprise data into automated action. Databricks, Snowflake's closest competitor, is making a similar push with its own AI and data platform strategy.
The key differentiator Snowflake is banking on is data gravity — the idea that because so much enterprise data already lives inside Snowflake, it has a natural advantage in deploying agents that need access to that data. Competitors building agents on top of third-party data stores face additional complexity around integration, governance, and latency.
Whether data gravity translates into workflow adoption remains to be seen. But Ramaswamy is betting that the company best positioned to win the agentic enterprise race is the one that already owns the data layer — and Snowflake intends to be that company.
What This Means for Enterprises
For enterprise decision-makers, the message from Snowflake's pivot is clear. The data warehouse is evolving into something far more ambitious — a platform that does not just store and analyse information but actively uses it to complete work. The companies that adapt to this shift early will gain a productivity advantage. Those that continue treating their data platforms as passive storage risk falling behind as competitors deploy AI agents that move faster and execute autonomously.
The era of simply warehousing data is ending. The era of putting it to work has begun.
