Upwind Raises $400M at $3.8B as Revenue Quadruples and a Datadog Bid Falls Away

Upwind, the runtime cloud and AI security company, closed a $400 million Series C at a $3.8 billion valuation led by Bessemer Venture Partners, the company said in its October 28 release. That is about 2.5x the $1.5 billion mark from its $250 million Series B in January, and Upwind says revenue has quadrupled in 2026.

The numbers

• $400M Series C at $3.8B, led by Bessemer; Canapi Ventures is the one new investor.
• Every existing backer returned: Greylock, Cyberstarts, Craft, Salesforce Ventures, TCV, Leaders Fund, Alta Park, Penny Jar, Cerca, Swish and Picture Capital, per CEO Amiram Shachar’s blog post.
• Over $800M raised since the company was founded in 2022.
• Less than 9 months between the Series B and the Series C.
• About $90M in ARR projected for year-end, growing $20M to $30M a quarter, according to a person familiar with the financials cited by TechCrunch. That puts the round at more than 40x forward ARR.

The Datadog detour

TechCrunch reports the round nearly fell apart when Datadog tried to buy Upwind. Datadog did not offer a premium over the $3.8 billion venture price, the outlet’s source said, so Upwind stayed independent. Neither company confirmed the talks in the official announcements. Shachar previously sold cloud cost startup Spot.io to NetApp for $450 million.

The product: security from the running workload out

Upwind’s pitch is runtime context. Instead of relying on periodic scans, it watches live cloud workloads, APIs and AI agents to decide which exposures are actually reachable and to catch attacks as they happen. It runs on AWS, Azure, Google Cloud and Oracle Cloud, and plugs into AWS Security Hub with billing through AWS. Named customers include Siemens, Roku, Carvana, Nubank, Supabase and ClickUp.

The AI layer is three agents. Blue investigates across log sources and assembles one attack story. Red safely tests whether a finding is exploitable in the customer’s environment. Green finds the fixes that clear thousands of findings at once. Upwind also launched an AI Labs group covering AI security, agentic security, frontier models and benchmarks, and advanced AI systems.

Where the money goes

Shachar says the capital goes to product and engineering: AI sensors on endpoints, scanners for agents and skills, a rebuilt inventory graph, and expansion into identity and data security. His framing is a move from “code to cloud” to “endpoint to code to cloud.”

What we are watching

Upwind discloses growth as a multiple, not a number. The ARR figure and the Datadog approach both come from an unnamed source, not the company. At more than 40x forward ARR, the $3.8 billion price assumes the runtime-plus-agents pitch beats Wiz-style posture tools now owned by Google and the platform bundles from CrowdStrike and Palo Alto Networks. The near-term proof is whether the agent products turn into paid seats, and whether the next disclosed revenue number keeps the 4x pace.