Cylake, a startup founded by Nir Zuk, the former chief technology officer of Palo Alto Networks, has secured $245 million in a convertible note to build a cybersecurity platform that operates entirely on premises without relying on cloud infrastructure. The funding round, led by Lightspeed Venture Partners alongside Picture Capital and Redpoint Ventures, brings the company’s total funding to $290 million since its 2026 launch.
The investment targets a specific but underserved market: organizations bound by regulations or security requirements that prohibit sending sensitive data to external cloud providers. Financial services firms, government agencies, and critical infrastructure operators cannot risk storing security information outside their jurisdictions or control. For these customers, modern cloud-first cybersecurity vendors have largely stopped offering viable options, forcing them to rely on aging on-premises products from vendors like Splunk, McAfee, and Symantec or to assemble fragmented systems from open-source components.
Cylake’s approach addresses this gap by building a unified, locally operated security system that gives organizations complete sovereignty over their infrastructure and data while still delivering AI-driven threat detection and response capabilities. The company plans to reach beta testing by the end of 2026 and launch commercially in 2027.
Why On-Premises Security Became Necessary
Traditional cybersecurity architecture assumes organizations can send data to the cloud for analysis and storage. That model fails for enterprises with strict data residency requirements. Banking regulations in Switzerland, for example, mandate that financial data remain within Swiss borders. Cloud-dependent security tools force these organizations into a difficult choice: compromise data sovereignty or operate with outdated tools.
Building hardware became central to Cylake’s design once the company determined its system had to operate independently while processing massive volumes of security data. Hardware ownership also gives Cylake control over cost and capacity, allowing customers to run AI models and store security data locally without paying separate cloud storage fees or token-based AI charges.
“We didn’t come from the perspective that we wanted to build hardware and find a purpose for it,” said René Bonvanie, Cylake’s chief marketing officer. “We wanted to build a system that could work on all of an organization’s data, a system that was not dependent on the cloud and it turns out the only way you can do that is when you also build the hardware.”
Consolidating Fragmented Security Tools
Most organizations rely on dozens of disconnected security tools, each collecting data from a different part of the infrastructure. Endpoints feed one system, servers feed another, and identity management runs on a third. This fragmentation creates blind spots that attackers routinely exploit.
Cylake plans to create a single data lake fed by a common ingestion pipeline, collecting information from endpoints, servers, IoT devices, and third-party security products. This unified approach allows security operators to analyze patterns across an organization’s complete dataset rather than switching between isolated tools and managing separate billing.
The company intends to provide endpoint security, data protection, IoT security, identity management, and browser security itself, gradually replacing the point products customers currently use. Third-party telemetry will remain available, but Cylake’s own technologies will generate increasingly more “firsthand” data over time.
Building AI Capabilities In-House
Implementing AI-driven threat analysis on premises creates a technical challenge: the system cannot simply send customer data to public AI providers like Anthropic or OpenAI without undermining the sovereign model at the platform’s core. Cylake must develop or adapt AI capabilities that run entirely within its own system.
The company plans to combine open-source AI models with proprietary technology tailored for enterprise security use, ensuring customers retain full control over how their data is processed. This requirement drove much of Cylake’s engineering roadmap and helps explain why the company needed $245 million now rather than spreading the capital across multiple funding rounds.
The funding will accelerate hiring in engineering and product divisions as Cylake prepares design partners for beta testing and eventual commercial deployment. The company has grown from 31 employees at the time of its March seed round to over 40 staff members, with further recruitment planned across core technical and operational areas.
A Market Shift Away From Cloud-Everything
This funding level suggests investors and enterprise customers are questioning the assumption that all security infrastructure should move to the cloud. For organizations handling sensitive data, regulatory compliance, or national security implications, that assumption has always been problematic. The gap between what the market offered and what regulated enterprises needed created an opening for a company willing to rebuild security architecture around premises-based deployment and data control.
Whether Cylake can execute on this vision depends on delivering both the unified data platform and the AI capabilities it promises by 2027. Success would challenge the prevailing industry model and offer a template for building enterprise infrastructure that balances modern technology with data sovereignty requirements.
