Nov 6, 2025 · 42m · no-priors
No Priors Ep. 139 | With Snowflake CEO Sridhar Ramaswamy
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this episode of No Priors, Snowflake CEO Sridhar Ramaswamy discusses leading the company's AI transformation, explaining how Snowflake Intelligence, multi-cloud neutrality, and disciplined organizational restructuring enable enterprises to achieve practical return on investment from AI.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. The hosts hold 16.9% of the talking time here. How this is scored →
The hosts as informed peer 5.4 Guest teaching 4.7 Guest disagreement 1.3 The hosts pushing back 1.5
speaking balance: gold is the hosts, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 39:40 Rejecting the maximalist LLM premise
Sridhar firmly rejects the idea that LLMs should handle everything internally, arguing that smart engineering relies on Python tools and search rather than brute-forcing calculations in model context.
Hardest push from the hosts ▶ 11:55 Sarah challenges platform boundary expansionSarah directly questions Sridhar on whether Snowflake's sales assistant and agentic platforms are overstepping the line from core data infrastructure into full-fledged applications.
Biggest teaching moment ▶ 38:48 Demystifying PageRank and search feedback loopsSridhar educates the audience on search history, explaining that Google's PageRank ran out of steam by 2004 and that real long-term platform value came from continuous user click feedback loops.
The host holds their own ▶ 22:45 Sarah synthesizes enterprise software defensibilitySarah articulates a sharp thesis on defensibility being continuously built through execution rather than abstract strategizing in an uncertain AI landscape.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | The hosts as informed peer | Guest teaching | Guest disagreement | The hosts pushing back | Why |
|---|---|---|---|---|---|---|
| Taking the Helm and Leading Snowflake's AI Transformation | 4 | 3 | 1 | 1 | Sarah sets the stage by asking Sridhar about his first 18 months as CEO after taking over from Frank Slootman. Sridhar provides a warm, reflective overview of navigating market doubts and steering Snowflake into an AI-first product focus. | |
| Restructuring Snowflake for Speed, Accountability, and Product Velocity | 5 | 4 | 1 | 1 | Sarah prompts on organizational prioritization and interjects on Snowflake's historical dominance as a pure data warehouse. Sridhar explains flattening engineering layers for speed and finding their identity as an AI Data Cloud. | |
| Announcing Snowflake Intelligence and Purpose-Built Enterprise Agents | 4 | 5 | 2 | 1 | Sridhar openly discusses halting their foundation model training due to capital reality and choosing to build an opinionated enterprise agent platform. He contrasts Snowflake Intelligence with broad, unfocused CSP agent platforms. | |
| Democratizing Data Access with Rigorous Evals and Consumption Pricing | 5 | 5 | 1 | 1 | Sarah clarifies whether Snowflake Intelligence is built for SQL practitioners or non-technical business users. Sridhar explains consumption pricing and the necessity of strict software engineering evals rather than YOLO AI. | |
| Navigating the Boundaries Between Data Platforms and Enterprise Software | 6 | 4 | 2 | 4 | Sarah pushes Sridhar on blurring lines between Snowflake's agent platform and SaaS applications, noting his known impatience. Sridhar acknowledges that the line between pure software and agentic data systems will be bloody. | |
| Lessons from Academia, Google Scale, and Founding Neeva | 4 | 4 | 1 | 1 | Sarah asks about Sridhar's personal journey from academia to Google scale and founding Neeva. Sridhar shares humble reflections on academic abstract writing, startup heartbreak, and Google's immediate distribution power. | |
| Building Defensibility in the Age of Hyperscalers and AI Labs | 7 | 5 | 2 | 2 | Sarah frames the strategic game theory of building on hyperscalers and frontier AI labs. Sridhar describes frontier labs as expanding empires that haven't met their oceans yet and warns against falling behind like Intel. | |
| The Multi-Year Vision for Snowflake as an Enterprise Data Companion | 5 | 6 | 1 | 1 | Sridhar outlines the multi-year vision of Snowflake as an enterprise data companion, contrasting 1990s slow Photoshop release cycles with real-time feedback loops in Google Search ads. | |
| Expanding Strategic Partnerships Across Cloud Providers and Enterprise Systems | 6 | 4 | 1 | 2 | Sarah probes partnerships with ecosystem giants like SAP and Microsoft, adding a witty quip on managing partnerships. Sridhar discusses transitioning to bidirectional data sharing and collaborative agent ecosystems. | |
| Stack-Ranking High-ROI AI Use Cases and Iterative Experimentation | 6 | 5 | 1 | 1 | Sarah asks for an enterprise ROI stack rank. Sridhar highlights coding agents and customer support, advising enterprises to iterate in thousand-dollar increments rather than placing massive unproven bets. | |
| The Future of Digital Advertising and Citations in Conversational AI | 6 | 4 | 1 | 1 | Sarah inquires about the fate of the digital advertising model in conversational AI interfaces. Sridhar asserts advertising will adapt while emphasizing user agency, and Sarah highlights the growing importance of primary source citations. | |
| Combining Information Retrieval, Evaluation Loops, and Specialized Tools | 7 | 7 | 2 | 2 | Sarah asks if traditional information retrieval and search indexing are becoming obsolete. Sridhar dismantles the maximalist AI view by explaining PageRank history, click feedback loops, and why models must call external tools rather than doing raw compute internally. |
Statements from this episode (18)
Ramaswamy: Slootman Stepped Down Because Snowflake Was Slow on AI
“I think what happened was the company was a little slow to reacting to changes from things like machine learning and AI. And that was a little bit of, honestly, the reason why Frank voluntarily pushed for the change, because he felt presently that we are heade…”
Ramaswamy: Hypergrowth created 7 to 10 organizational layers at Snowflake
“Snowflake had basically specialized at every layer possible, and there was a very long distance between the engineer that did a feature and the customer that made use of the feature, and there were like seven to 10 layers of teams that were involved.”
Ramaswamy: Snowflake Abandoned Foundation Models Due to Capital Constraints
“Early last year, we actually went down the path of creating foundation models. We created a credible MOE model. This was early last year, but we also quickly realized that our ability to compete with the likes of OpenAI or Anthropic was going to be really hard…”
Ramaswamy: 2D Dashboards Fail to Answer Complex Enterprise Questions
“A dashboard is a two D view of a complex surface. It just has no easy answers to the many questions that any reasonable person you or I is going to have off of that.”
Ramaswamy: Enterprise AI requires rigorous evals like software engineering
“We need to think of AI the same way we think about software engineering, which is there's a right and there's a wrong. It cannot be this mode of like, YOLO, AI, you can get some good answers, some terrible answers, it's your problem. And so we very much emphas…”
Ramaswamy: Snowflake is not trying to replace SAP or Salesforce
“On the other hand, it's absurd if we think we are SAP or Salesforce. We are not. There are, you know, somebody managing a hundred billion dollar supply chain ecosystem with a complicated software provider isn't saying, hey, I'm going to use SI and I don't need…”
Ramaswamy: Line Between Agentic Systems and Pure Software Will Be Bloody
“But on the other hand, I think the line between what an agentic system like this is going to be and what like pure software is going to be will be absolutely is going to be bloody.”
Ramaswamy: Snowflake deployed coding agents to all solution engineers
“We've rolled out coding agents to all of our solution engineers, and they are excited because that just dramatically lowered the amount of time it takes to create a demo.”
Ramaswamy: OpenAI and Anthropic are laser-focused on building top coding agents
“Coding agents are particularly interesting from this perspective because it is very clear that both Anthropic and OpenAI are going to be laser set on having the best one best one that there is.”
Ramaswamy: Why Google Failed in Shopping and Travel
“Google, for example, stopped that information. God knows I spent enough time trying to get into physical things like shopping or airline purchases or hotels. We didn't really succeed because we didn't have core competence really in some fundamental way beyond …”
Ramaswamy: Tech Companies That Do Not Innovate Will Become Intel
“Unless you innovate and stay ahead, not just be ahead, but stay ahead, you will be Intel.”
Ramaswamy: Google and Meta Succeeded by Being Data-First Rather Than Product-First
“The great companies of this century, a company like a Google or a Meta, were more data-first companies than like purely product-first companies than pretty much any others compared to before.”
Ramaswamy: A Multi-Cloud Data Platform Can Become as Large as a CSP
“A data platform that especially spans CSPs has the right, you have to earn it, has the right to be as large as a CSP itself.”
Ramaswamy: Databricks and Fabric Strained Snowflake's Relationship With Microsoft
“Among the earliest places where this went to work was in our relationship with Microsoft, which was okay, but not that great because they had a one P relationship with Databricks. And they're always kind of conflicted about is fabric the answer or snowflake th…”
Ramaswamy: Coding Agents Offer Easiest Enterprise AI ROI
“I would say that coding agents are probably the easiest auto I just in terms of making new projects faster, making tech demystifying technology so that more people can get at it.”
Ramaswamy: Advertising Will Reinvent Itself in the Conversational AI Era
“It will reinvent itself in the chat world.”
Ramaswamy: Google's PageRank Ran Out of Juice by 2004
“The insight that powered Google was PageRank. It was a way of harnessing the power of the entire internet to figure out what was popular and what was not. But PageRank ran out of juice. Like in six years, like 2004 or five. And while Google never really liked …”
Ramaswamy: AI agents should use specialized tools over maximalist LLMs
“A smarter person is going to say, no, they should not do math. Instead, I should write the two lines of Python, which I know how to, you know, direct and run the Python in order to solve the math problem. I think of trust in a very similar way. There are well-…”