Upwind has raised $400 million in a Series C round at a $3.8 billion valuation, less than nine months after its previous financing valued the company at $1.5 billion. The San Francisco startup also turned away a Datadog acquisition approach, after Datadog declined to pay more than the new valuation.
Bessemer Venture Partners led the round, with existing investors Greylock Partners, Cyberstarts, Craft Ventures, Salesforce Ventures and TCV taking part. Canapi Ventures joined as a new investor. The round takes Upwind’s total funding past $800 million since it was founded in 2022.
A buyout offer that fell short
The financing had been in progress for months when Datadog approached Upwind about a purchase. Datadog sells software that monitors how applications run, but it would not pay more than the $3.8 billion valuation set by venture investors. Upwind chose to stay independent instead.
Upwind was founded by Amiram Shachar, who previously built Spot.io and sold it to NetApp for $450 million in 2020. The company says its revenue quadrupled during 2026, and it is projected to end the year with annual recurring revenue of about $90 million.
What the money will fund
Upwind plans to put the capital into product and engineering growth. It is building AI sensors for endpoints, scanners for AI agents and skills, and a rebuilt inventory graph that maps the relationships between a customer’s assets. It also plans to expand into identity, data and AI security.
The company sells a runtime security platform that watches cloud workloads, application programming interfaces and AI agents while they run. Instead of relying on periodic scans, it uses live data to find real threats and cut through alert noise. Customers named by the company include Siemens, Roku and Carvana.
“Security is a cloud and AI story now,” said Amiram Shachar, co-founder and chief executive of Upwind.
Agents that answer AI attacks
Upwind argues that AI attacks must be answered with AI. Its Blue agent pulls logs and data points into a single picture of an attack. Its Red agent checks whether a weakness can actually be exploited. Its Green agent finds fixes that clear thousands of findings at once.
The company recently launched an AI Labs research unit focused on threats to AI systems and agentic workflows. Upwind says the lab gives researchers room to work on security problems that will define what the industry needs next.
