Neo

Neo is a cybersecurity company focused on securing AI agents and agentic applications within enterprise environments. The platform is designed to give security operations (SecOps) teams visibility, attribution, and control over all agentic software activity across endpoints in real time. The core platform consists of four functional layers: - Discovery: Neo deploys a lightweight endpoint sensor (approximately 25MB, requiring no kernel module or system reboot) that creates a live inventory of all software on managed and unmanaged endpoints, including AI agents, MCP (Model Context Protocol) servers, browser extensions, scripts, and plugins. - Configuration analysis: Neo inspects the configuration of discovered software to identify misconfigurations, excessive permissions, and security gaps — for example, agents with over-broad credentials or those configured to operate without human supervision. - Runtime monitoring: Neo monitors agent actions in real time as they interact with connected systems such as CRMs, databases, cloud services, code repositories, and collaboration tools. - Enforcement: Neo intercepts agent actions at the endpoint and applies policy decisions — allowing, blocking, or holding actions for human approval — before sensitive data or systems are accessed. Policies can be defined using natural language queries processed by a built-in LLM. Neo also includes a knowledge base called Neoverse, which continuously maps the capabilities, risks, and behaviors of known agentic software (e.g., Claude Code, GitHub Copilot, Cursor, Devin) and is bundled with every deployment. The platform targets enterprise security teams and is compliant with SOC 2 Type II, GDPR, and HIPAA. Neo is backed by investors including Andreessen Horowitz, Bessemer Venture Partners, Craft Ventures, and Merlin Ventures.