{
  "schema_version": "1.1",
  "id": "archive:https://openai.com/index/jiostar-vijay-seshadri",
  "slug": "jiostar-vijay-seshadri-0nt0fel",
  "url": "https://feed7.dev/p/jiostar-vijay-seshadri-0nt0fel",
  "title": "JioStar’s chief architect on AI discovery at 500M-user scale",
  "why_included": "JioStar’s conversational discovery launch highlights the work coding-agent teams need around evals, guardrails, domain experts, and production monitoring at scale.",
  "summary": "JioStar built a conversational video and cricket-stat discovery assistant for JioHotstar in about **eight weeks** and across **more than 15 languages**. Its platforms serve over **500 million consumers** and have supported up to **72.5 million concurrent viewers**.\nFor builders running coding agents, the interview stresses defining what the product should not answer, using domain experts to seed evals, and feeding production failures into new test cases. Codex sped up some implementation and eval work, while people remained responsible for validation.\nThe material gives no measured model-efficacy or latency results for the assistant. The scale figures describe JioStar’s platforms, not necessarily traffic handled by the new feature.",
  "practical_implication": "",
  "agent_context": "JioStar built a conversational video and cricket-stat discovery assistant for JioHotstar in about **eight weeks** and across **more than 15 languages**. Its platforms serve over **500 million consumers** and have supported up to **72.5 million concurrent viewers**.\nFor builders running coding agents, the interview stresses defining what the product should not answer, using domain experts to seed evals, and feeding production failures into new test cases. Codex sped up some implementation and eval work, while people remained responsible for validation.\nThe material gives no measured model-efficacy or latency results for the assistant. The scale figures describe JioStar’s platforms, not necessarily traffic handled by the new feature.",
  "source": {
    "name": "OpenAI",
    "url": "https://openai.com/index/jiostar-vijay-seshadri",
    "published_at": "2026-11-11T00:00:00.000Z"
  },
  "source_class": "blog_post",
  "content_type": "Official Release",
  "layer": "benchmark",
  "domains": [
    "coding",
    "data"
  ],
  "topics": [
    "agent-evals",
    "agent-reliability",
    "harness-engineering"
  ],
  "verification": {
    "status": "official_source",
    "label": "Official Source",
    "method": "source_feed",
    "verified_at": null
  },
  "uncertainty": [],
  "connected_context": {
    "meaning": "This grounds agent-evaluation guidance in a multilingual product built for a very large platform, while clarifying that platform scale does not establish traffic or model quality for the new assistant. Its strongest contribution is an operating loop: define answer boundaries, seed tests with domain experts, and turn production failures into eval cases while people validate the results.",
    "corpus_size": 1067,
    "generated_at": "2026-11-12T11:01:04.666Z",
    "connections": [
      {
        "title": "From Agent Traces to Agent Simulations — Rustem Feyzkhanov, Snorkel AI",
        "source_name": "AI Engineer",
        "source_url": "https://www.youtube.com/watch?v=Ib5t2RLtxvM",
        "feed7_url": "https://feed7.dev/p/from-agent-traces-to-agent-simulations-rustem-feyzkhanov-snorkel-ai-0zwlzjq",
        "reason": "JioStar describes converting production failures into tests; trace replay adds a way to compare agent stacks against recorded cases under fixed conditions."
      },
      {
        "title": "Your Agent Evolved. Your Evals Didn't. — Ameya Bhatawdekar, Braintrust",
        "source_name": "AI Engineer",
        "source_url": "https://www.youtube.com/watch?v=nxokqOq1imY",
        "feed7_url": "https://feed7.dev/p/your-agent-evolved-your-evals-didn-t-ameya-bhatawdekar-braintrust-1loaqv2",
        "reason": "The production-failure loop aligns with architecture-aware evals: as the assistant’s tools or orchestration change, its tests need to cover the changed system, not only answer quality."
      },
      {
        "title": "Building GTM AI Agents: Lessons from Deploying to 6,000 Users — Sait Izmit, Snowflake",
        "source_name": "AI Engineer",
        "source_url": "https://www.youtube.com/watch?v=DrTdD-ttjCY",
        "feed7_url": "https://feed7.dev/p/building-gtm-ai-agents-lessons-from-deploying-to-6-000-users-sait-izmit-156ari4",
        "reason": "Both emphasize narrow, deliberate coverage and production feedback; Snowflake adds a staged expansion pattern, while JioStar stresses domain-expert test seeding and explicit answer boundaries."
      }
    ]
  },
  "lifecycle": "Current",
  "published_at": "2026-11-11T00:00:00.000Z",
  "modified_at": "2026-11-11T00:00:00.000Z",
  "supersedes": [],
  "expires_at": null,
  "formats": {
    "html": "https://feed7.dev/p/jiostar-vijay-seshadri-0nt0fel",
    "json": "https://feed7.dev/p/jiostar-vijay-seshadri-0nt0fel.json",
    "markdown": "https://feed7.dev/p/jiostar-vijay-seshadri-0nt0fel.md"
  }
}