Key Takeaways
- • Lightspeed Venture Partners is leading a $245M financing for Cylake, a security startup, according to a Lightspeed-published investment announcement.
- • Cylake is led by Nir Zuk, co-founder of Palo Alto Networks, with Wilson Xu (former Palo Alto Networks engineering head) and Udi Shamir (SentinelOne co-founder) on the founding team.
- • Cylake proposes consolidating the typical enterprise stack of 30-plus security vendors into a single on-premises or managed-service "private cloud" data lake, targeting regulated industries that cannot send data to the cloud.
- • The company's pitch centers on customer-owned appliances that reduce the marginal cost of AI-driven security analytics, contrasting with per-token cloud AI pricing.
Summary
The founding team also includes Wilson Xu, who ran engineering at Palo Alto Networks, and Udi Shamir, co-founder of SentinelOne. Lightspeed characterizes the group as having contributed heavily to the modern security stack and says all three share the view that security requires a new architecture.
Lightspeed frames the problem as fragmentation. Large enterprises often run more than 30 security vendors, creating separate data silos, dashboards and bills, and security teams spend substantial time integrating tools. Regulated sectors such as banks, governments and critical infrastructure operators face additional constraints: many modern cloud-based platforms cannot be used because data cannot leave the organization for compliance or sovereignty reasons, leaving some of these organizations on tools more than two decades old. Lightspeed says the problem has grown for years and that AI has made it more urgent as attacks become harder to detect and security data volumes exceed what human teams can monitor.
According to Lightspeed, Cylake delivers a complete appliance-based security stack inside the customer's environment, either in its own data center or in facilities run by a managed service provider. All security data streams into a single lake, and security functions run as applications on that shared data. Machine learning monitors the data in real time to flag anomalies, and large language models investigate findings, explain incidents and help build remediation alongside security teams. Lightspeed argues that because the appliances are customer-owned, the marginal cost of AI analysis is close to zero, in contrast to cloud architectures that charge per AI token.
The primary source is a firm-authored investment announcement and does not disclose valuation, other investors, or product availability timelines. Figures beyond the $245 million round were not provided in the source material.
Go deeper than the headline
Research every firm in this story, track the ones that matter and get the daily briefing — free for 7 days.