Upwind raises $300M Series C at $3.8B valuation for cloud security

Upwind closes $300M round eight months after last raise

Upwind Security has secured $300 million in Series C funding, bringing its valuation to $3.8 billion — a $2 billion increase since the beginning of the year. The round was led by returning investors Bessemer and TCV, with participation from Salesforce Ventures, Greylock, Craft Ventures, and other firms.

The funding comes just eight months after Upwind's previous nine-figure round, signaling strong investor confidence in the company's approach to automated cloud security.

Why it matters

Cloud environments present unique security challenges because they change constantly as developers add and remove workloads, instances, and user accounts. Traditional security tools struggle to keep pace with these rapid changes, creating gaps that attackers can exploit. Upwind's ability to refresh its asset mapping every 30 seconds addresses a critical pain point for enterprises managing complex, dynamic cloud infrastructures.

Technical approach using eBPF

Upwind's platform automatically maps all assets within a cloud environment, including instances, encryption keys, configuration scripts, and their interconnections. The system updates this map every 30 seconds to capture configuration changes in real time.

The company's technical foundation relies on eBPF, a Linux kernel feature that enables data collection without making error-prone modifications to the operating system. eBPF deploys observability code in an isolated sandbox, preventing bugs from affecting sensitive kernel modules while still gathering comprehensive telemetry about workloads.

This approach allows Upwind to collect detailed information about cloud assets and their relationships — for example, showing administrators exactly which documents a data visualization application can access.

AI-powered vulnerability detection

Upwind employs artificial intelligence to scan discovered cloud assets for security weaknesses. The platform compares workload configurations against cybersecurity best practices and regulatory requirements, then launches simulated cyberattacks against critical assets to test their defenses.

The company uses two AI agents to prioritize threats: one called Red filters out low-risk vulnerabilities that are unlikely to lead to actual breaches, while another named Blue identifies active hacking attempts.

The platform also includes specialized features for monitoring AI models, automatically generating AI-BOMs (bills of materials) that list an AI application's components, making it easier to identify risky software dependencies.

Growth trajectory

While Upwind hasn't disclosed specific plans for the new capital, recent activity suggests a focus on ecosystem expansion. Over the past month, the company has launched integrations with approximately six third-party cloud services, indicating a strategy centered on broadening platform compatibility.

The details were first reported by CTech and SiliconANGLE.