It's the data query engine, stupid

Welcome to Runtime! Today: Snowflake's massive quarter shows once again that AI deployments hinge on data quality, Anthropic drops Opus 4.8 and picks up an enormous pile of cash, and the latest enterprise moves.

Please forward this email to a friend or colleague! If it was forwarded to you, sign up here to get Runtime each week, and if you value independent enterprise tech journalism, click the button below and become a Runtime supporter today.

Sorry, 'stupid' was a little harsh

If it hasn't already been used to death, let's finish off that well-worn phrase right here: When you find yourself in a gold rush, figure out how to sell the picks and shovels. The companies that saw the earliest success from generative AI tools were the companies that had a strong data foundation already in place, and the long-awaited arrival of AI agents has the rest of enterprise tech committing to a data overhaul.

Snowflake just turned in one of the best quarters in its history as a public company, reporting accelerating revenue thanks to enterprise investment in its data and AI tools. "The combination of Snowflake's trusted enterprise data, rich business context, leading AI models, and secure connectivity into enterprise applications creates a unique opportunity" to keep that momentum going, said CEO Sridhar Ramaswamy on a conference call Wednesday following the results.

• Revenue came in at $1.3 billion, up 34% from a year ago amid the "strongest sequential dollar growth in company history," Ramaswamy said.
• Snowflake also raised its guidance for the full year, now projecting that it will grow revenue by 31% this year compared to last year, which sent its stock soaring nearly 37% on Thursday.
• The company is seeing increased demand for its core products as customers move their older data into its core platform to connect AI applications: "The number of use cases, individual projects managed on Snowflake deployed in the quarter increased 114% year-over-year as customers moved more workloads into production on the platform," he said.

Contrast that performance with the quarter turned in by Salesforce, which beat Wall Street's expectations for the period but signaled Wednesday that already-depressed expectations for future growth would again need a readjustment. Only Marc Benioff could open an earnings call by saying "agentic AI, well, it's the biggest growth opportunity for our customers," minutes before CFO Robin Washington predicted Salesforce revenue growth would slow to 10% during the current quarter.

• Salesforce's problem is that while it manages incredibly important data on behalf of its customers, those customers appear to prefer using external tools like Snowflake or Databricks to ready that data for consumption by AI agents.
• That's just one reason why Salesforce has embraced the "headless" strategy for its core services, presenting a new MCP interface to agents and other tools that don't need to log into Salesforce's UI through a browser.
• Salesforce has long offered up its services through an API, but "the Headless 360 strategy that Marc walked through expands our addressable market into surfaces we've never previously monetized," Washington said.
• The company cited Anthropic as an example of how the headless strategy could reinvigorate Salesforce growth; according to president and chief revenue officer Miguel Milano, Anthropic's use of Sales Cloud "exploded fivefold" after tapping into the headless service with internal agents.

Companies still need Salesforce data to run their sales and marketing operations, but they need Snowflake (and Databricks) to help them get all that data ready for AI agents, as well as the data that runs the rest of their business. That trend benefits Snowflake more than Salesforce, especially as companies continue to embrace data lakes and lakehouses that let businesses combine data into a single place.

• "Across industries, organizations are moving toward a future where employees and intelligent agents work side by side to accelerate decisions, automate complex workflows, and unlock entirely new levels of productivity and innovation," Ramaswamy said.
• There's clearly a role for Salesforce in that vision — especially as a headless service — but it's a secondary, less central role than it is used to playing in enterprise tech.
• "What's happened this year is initially the year started with all software companies going down. And what's happened in the last few weeks is we're seeing a separation between winners and losers," said D.A. Davidson's Gil Luria on Bloomberg TV.

H is for heaping

Anthropic positioned itself as the enterprise AI model maker long before OpenAI decided to follow suit last year, and on Thursday it announced two big developments that will further those ambitions. Armed with billions in fresh cash and a new edition of its flagship Opus model, Anthropic will look to build on a $47 billion annual revenue run-rate right as enterprise AI is starting to take hold.

Anthropic's $65 billion Series H funding round values the company at $965 billion, which vaults it ahead of OpenAI after nearly tripling the valuation delivered by its last funding round in February. "This latest funding is expected to advance our safety and interpretability research, expand compute to meet growing demand for Claude, and scale the products and partnerships our customers rely on," the company said in a blog post.

The company also released Opus 4.8, which becomes the top-of-the-line model in Anthropic's arsenal but should be considered "a modest but tangible improvement on its predecessor," the company said. It also sounds like a much wider group of people are about to get their hands on Mythos Preview: "We’re making swift progress on developing these safeguards and expect to be able to bring Mythos-class models to all our customers in the coming weeks," it said.

Enterprise moves

Ashraf Alkarmi is the new co-CEO of Dropbox, and will replace co-founder and current CEO Drew Houston after a transition period.

Jean-Michel Lemieux is the new executive individual contributor at Spellbound, joining the legal tech company in a newly created role where he will "[function] as a high-leverage operator working across product, engineering, go-to-market, and internal systems for the AI-native enterprise."

Daniel Raskin is the new chief marketing officer at Virtana, joining the observability company after marketing leadership roles at Mperativ and Kinetica.

Colin Fleming is the new chief marketing officer, business, at OpenAI after marketing leadership roles at ServiceNow and Salesforce.

The Runtime roundup

In addition to its blowout earnings, Snowflake announced Wednesday that it has acquired MCP specialist Natoma and signed a $6 billion compute deal with AWS.

Dell reported staggering first-quarter revenue growth of 88% compared to the last year, lifted by a 757% increase in sales of AI servers.

Meta would consider selling excess compute capacity if its AI plans fail to materialize, according to Mark Zuckerberg, which would see it enter a market the company has been toying with for years.

The brief "tokenmaxxing" era appears to be dead after Amazon ditched an internal AI leaderboard that incentivized all the wrong things, according to The Financial Times.

Thanks for reading — see you Saturday!