Comp AI sets eyes on a continuously agentic future for security and compliance

Comp AI announced a $34 million Series A financing round led by Roo Capital and Grand Ventures as it ramps up a platform aimed at automating security and compliance tasks in environments that increasingly use AI agents. The startup was formed in January by Lewis Carhart and brothers Claudio and Mariano Fuentes; the trio previously worked together on a workflow platform called LeapAI, which operated for roughly two years and scaled to more than one million users before the team shut it down. The founders say their prior experience highlighted two lessons: the technical challenges of building with large language models and the operational burden of compliance processes such as SOC 2. That burden, they argue, often pulls engineering attention away from product work when teams manually assemble evidence and documentation for audits, a gap Comp AI intends to address. Comp AI’s product is described as an agentic platform that assists with drafting security policies, collecting audit evidence, and continuously monitoring compliance controls. The company also offers AI-driven penetration testing that proactively scans codebases and infrastructure for vulnerabilities. The team emphasizes the product augments rather than replaces human roles: people onboard and supervise agents, approve policy drafts, and support controls as agent responsibilities grow. The startup positions itself among a growing cohort of cybersecurity firms focused on AI-era risks, citing situations where a completed audit can quickly become out of date if a company deploys new agents that access data or change permissions. Comp AI says it is starting with permissions and accountability features to record agent access and actions and to validate ongoing risk, and intends to use the Series A proceeds to expand its product capabilities.