Neo

Neo is an agentic software control platform designed to give security operations teams visibility and policy enforcement over AI tools and autonomous software running within an organization. The platform operates through a lightweight endpoint agent (25MB) that observes application behavior, connects it to identity and intent, and enforces policy at runtime. It is designed to complement existing endpoint, identity, network, and data security controls rather than replace them. Core capabilities include: - Inventory and discovery: Neo scans and catalogs all AI tools, agents, extensions, MCP servers, CLIs, and embedded AI functions across endpoints. An initial inventory scan is designed to complete in under 15 minutes. - Policy engine: Security teams can write and deploy policies using a built-in LLM chatbot interface. Policies can allow, block, redirect, or hold actions for human approval, and all decisions are preserved in an audit trail. - Runtime enforcement: Neo enforces policies at the application runtime layer, covering actions such as blocking unsigned MCP servers, quarantining prompt injection attempts, restricting file access, flagging credential exposure, and holding large data exports. - Audit mode: Teams can begin in a non-enforcement audit mode to inventory activity and validate policy impact before activating enforcement rules. - Identity and SOC integration: Full deployment is designed to integrate with identity providers (IdP) and SOC environments within one hour. Neo targets AI tools including coding assistants, desktop AI applications, and AI-embedded SaaS platforms. It addresses risks arising from agentic software that can interpret goals, invoke tools, chain actions, and operate autonomously under valid permissions.