Daxa

Daxa governs and secures AI agents at the point where they touch enterprise data. Its platform, built around the Pebblo data-classification engine and a Proxima AI retriever, inspects and tags sensitive data before it reaches a model or an autonomous agent, then applies real-time access control and behavioral governance so an AI system can’t act on data or take actions its policy doesn’t allow. Rather than securing an AI application from the outside, Daxa embeds controls into the data and retrieval layer that agents actually use.

The company was originally built as CloudDefense.io, a cloud-native application protection platform (CNAPP) covering code, container, and infrastructure security, before rebranding to Daxa and pivoting toward AI agent governance as generative AI deployment became the more urgent enterprise problem for its founding team. Founded in 2019 and based in Cupertino, California, Daxa was started by Huseni Saboowala and Sridhar Ramaswamy along with security veterans from Cisco, McAfee, F5, and Trend Micro, and raised a $3.13 million Series A in 2022. Daxa’s technology now runs inside HPE Private Cloud AI, governing both HPE’s own internally-deployed agents and those of HPE’s customers, and the company was included in Gartner’s 2025 Market Guide for AI Trust, Risk and Security Management (AI TRiSM).

Daxa’s pivot from CNAPP to AI governance reflects a broader shift among cloud security vendors racing to secure agentic AI deployments before they proliferate unmanaged; its HPE integration and Gartner AI TRiSM inclusion are real, named validation points, though the company’s Series A funding size suggests it remains an early-stage bet relative to larger, better-capitalized AI security entrants.