Aurva, the first unified platform for Access monitoring and AI observability, has launched out of stealth with $2.2 million in seed funding. The oversubscribed round was led by Nexus Venture Partners, with participation from DeVC and industry leaders including former Meta executives Chris Bream, Rahul Sood, Karandeep Anand, Mala Ramakrishnan and Postman founders Ankit Sobti and Abhinav Asthana.
Aurva was founded by former Meta engineers Apurv Garg (CEO) and Krishna Bagadia (CTO). Bagadia previously helped design and scale Meta’s internal data security systems, while Garg drove AI initiatives across business messaging units. Aurva’s approach is inspired by Hipster, Meta’s real-time data access control platform, and applies those lessons to high-scale enterprises now facing the security concerns from deploying AI-driven workloads.
Read More – Torus Raises $200M in Funding
From Static Access Control to Real-Time Data Usage Monitoring
From copilots embedded in tools to autonomous services operating in the background, AI-driven systems are accessing sensitive enterprise data in ways traditional tools can’t see or control.
Aurva gives enterprises end-to-end visibility into who is accessing sensitive data, how, and why, tying every query or data flow to a real identity, whether human, service, or AI agent.
“Existing tools were built for static environments. But AI is dynamic, it’s everywhere, and it just runs,” said Apurv Garg, founder and CEO of Aurva. “There are now countless ways data gets accessed, often without visibility. Enterprises need to rethink how they monitor sensitive data usage in real time; across humans, services, and AI agents. That’s why we built Aurva: to bring observability to AI usage and connect it directly to data access and flows.”
Legacy Database Activity Monitoring (DAM) tools were built for static environments, relying on delayed logs, generic alerts, and limited visibility into how data is actually used. In contrast, Aurva takes a fundamentally different approach, leveraging eBPF – a popular sandboxing approach at the kernel level used by Google, Netflix, Meta, and other enterprises to control software security – for real-time, low-overhead monitoring. The platform unifies query intelligence, agentic access monitoring, AI observability, and egress detection, helping enterprises detect overprivileged access, unauthorized AI usage, and risky data flows across services.
Read More – Aven Raises $110M in Series E Funding at $2.2B Valuation
