Snowflake wants to be the data workflow company

Welcome to Runtime! Today: Snowflake introduces new tools that promise to help companies find nuggets of insight in their corporate data, Thoma Bravo gears up for the coming wave of AI startup buyouts, and the latest funding rounds in enterprise tech.

(Was this email forwarded to you? Sign up here to get Runtime each week.)

Ride the pipeline

SAN FRANCISCO – Ask any sport-coat-and-jeans-wearing professional walking through downtown San Francisco this week to define their most precious corporate asset, and they'll probably say data. But working with data, especially the massive, sprawling data sets that are lying around the average enterprise, has long required companies to hire and train data scientists and engineers familiar with the complex procedures needed to extract value from that data.

Snowflake introduced two interesting generative AI services this week at its annual Data Summit that could make it easier to bring data sets together and expand the number of people who can work with them. The services are called Snowflake Openflow and Snowflake Intelligence, and in a press conference Monday CEO Sridhar Ramaswamy described them as an evolution of Snowflake's product development strategy to focus on the "future of workflows."

Snowflake Openflow emerged from the company's acquisition of Datavolo last year, executive vice president Christian Kleinerman said at the press conference.

• "Openflow is really the first time that we're providing connectivity to a massive number of different sources of data and allowing you to bring those easily and seamlessly into Snowflake," said Chris Child, vice president of product management, in an interview with Runtime.
• The managed service is based on Apache NiFi, an open-source project developed by the founders of Datavolo that provides a system for connecting data pipelines.
• It was designed to work with several types of data, including both structured and unstructured data as well as steaming data, and connectors are available to move data into Snowflake from databases like PostgreSQL, MySQL, and SQL Server or SaaS applications like HubSpot, Slack, and Workday.
• That allows companies to develop "workflows that can then be created with search and indices and models on top of that of the data," Ramaswamy said.

Think of Snowflake Intelligence as the data nerd's answer to the "vibe coding" movement. It allows employees who have questions about some aspect of their business to ask detailed questions in natural language about their data, expanding the pool of people who can work with that data.

• Snowflake Director of Product Jeff Hollan outlined a scenario in an interview with Runtime: A product manager sees a spike in revenue from a product dashboard, but is curious what sparked that increase.
• They will be able to ask Snowflake Intelligence to search across structured and unstructured corporate data related to that product without having to ask someone on the data team to build a query for them, which can take several hours.
• It's "putting an AI agent on top of all that enterprise data, and making it so that ChatGPT-on-top-of-enterprise-data [scenario] becomes real," Hollan said.
• Snowflake also introduced a service called AISQL that will allow data analysts to use SQL to work with unstructured data, which could help make it easier to build generative AI apps for internal or external use.

Snowflake also announced the $250 million acquisition of Crunchy Data on the first day of the conference, adding a PostgreSQL database company to its arsenal. Postgres is one of the most widely used databases in the world, and Snowflake intends to release its own managed service called Snowflake Postgres to work alongside its analytics tools.

• "We think this is an important acquisition in helping app developers, in helping users of Snowflake have access to a transactional store right within Snowflake," Ramaswamy said.
• The deal comes just weeks after rival Databricks paid $1 billion for Neon, a database startup that was also working with Postgres.
• "Crunchy gives Snowflake an enterprise-grade PostgreSQL engine to support AI agents, co-pilot apps, and context-aware workflows that demand structured, compliant, and low-latency operational data storage," said Constellation Research's Michael Ni, underscoring Ramaswamy's focus on workflows.

As the early hype around generative AI starts to thankfully settle down, companies are less interested in throwing it at everything they do and more interested in finding the more-limited but powerful ways it can automate the parts of their businesses that slow them down. Every company in enterprise tech has realized that this is a once-in-a-generation chance to reset the power rankings, and the ones who deliver are poised for growth.

• "No one feature that we create is difficult in and of itself, but it's the combination of what we represent as a data platform that we think is our unique and redeeming value," Ramaswamy said.

No owners, only spenders

Private equity has played an outsized role in enterprise software over the last decade, and Thoma Bravo is probably the highest-profile firm looking for companies that have promising technology but limited prospects. The rush of venture-capital funding into enterprise AI startups over the last couple of years and the continued slowdown in the IPO pipeline is going to create a lot of demand for exit strategies as the contenders and pretenders sort themselves out, and Thoma Bravo just reloaded.

The company announced Tuesday that it has raised $34.4 billion for three different funds that focus on software companies, according to the Wall Street Journal. Other private equity companies are struggling in the post-ZIRP era, but Thoma Bravo has enjoyed a 35% annualized return on its investments, according to the report.

Being acquired by a private equity company is generally not an outcome that favors customers or employees, but those transactions are going to happen as early shareholders look for returns on companies that need more funding than they are willing to provide. A lot of companies are going to find themselves in that situation over the next couple of years as AI strategies mature, and Snowflake and Databricks won't be able to buy them all.

Enterprise funding

Hex raised $60 million in Series C funding for its data-management and analytics platform.

CloudZero scored $56 million in Series C funding as it builds out its cloud cost-managment service.

Zero Networks landed $55 million in Series C funding for its security technology, which uses "microsegmentation" to help defenders stop lateral movement across their networks after a breach.

Chalk raised $50 million in Series A funding as it builds out data computing infrastructure for real-time inference.

Cerby scored $40 million in Series B funding for its automated identity security platform.

Ciroos launched with $27 million in Series A funding and announced its flagship product, which is a "SRE Teammate" that promises to help sort observability alerts.

The Runtime roundup

HPE beat Wall Street expectations for revenue and profit and raised its guidance for the current quarter, citing better-than-expected sales of AI servers.

Microsoft and CrowdStrike promised to come up with easier-to-follow standards for describing nation-state hacking groups, after realizing that previous attempts to make names like "Salt Typhoon" or "UNC2452" resonate weren't working.

Thanks for reading — see you Thursday!