Vinoth Govindarajan

Member of Technical Staff

Open AI

Vinoth Govindarajan is a Member of Technical Staff at OpenAI, where he works on core data infrastructure for large-scale AI systems and internal agent-facing platforms. His recent work includes internal facing agents, a support and on-call assistant for Data Platform that applies agent memory and retrieval to operational triage. He is co-author of Engineering Lakehouses with Open Table Formats and writes The Agent Stack, a systems-first publication on production AI agents and data infrastructure. His work focuses on control planes, state, memory, tool boundaries, reliability, and the architecture patterns that make agent systems safe and predictable in production.

Sessions (1)

Your Agent Didn’t Fail. Your Harness Did.

11:10 AM·Track 1 · Room 2010

AI agents do not fail only because the model is wrong. Many production failures happen in the harness around the model: state is not persisted, two runs mutate the same session, a tool call never returns, an approval loses scope, or an internal success never becomes user-visible proof. This talk uses OpenClaw as a public case study to examine real harness failure modes and extract a reusable production model for AI engineers. We will look at how events enter an agent system, how session state is rehydrated, why single-writer lanes and throttles matter, and why tool execution needs scoped approvals and auditable receipts. The core idea is simple: a model proposes, the harness commits, and the receipt proves it. Attendees will leave with a practical 'run receipt' audit they can apply to their own agents: what woke it up, which state did it inherit, what authority did it use, what executed, and what evidence survived.

Claws & Personal Agentsintermediatetalk