Sundar Vedaraman

Sundar Vedaraman

Helping teams govern AI runtimes and multi-agent infrastructure at Docker

Sundar Vedaraman works with engineering teams at Docker to solve the trust and infrastructure problems that show up when running AI agents in production.

As teams wire up multiple models across tools like Claude Code, Copilot, and Cursor, keeping track of permissions and guardrails gets messy fast. Sundar focuses on governing the runtime layer beneath those agents—controlling execution, tool calls, credentials, and token spend with strict default-deny policies so teams don’t have to rely on an agent making the right call.