{
  "schema_version": "1.1",
  "id": "s8:https://www.youtube.com/watch?v=jQDXzEVHMSE",
  "slug": "building-turbopuffer-gergely-orosz-pragmaticengineer-simon-eskildsen-ceo-04u1ucc",
  "url": "https://feed7.dev/p/building-turbopuffer-gergely-orosz-pragmaticengineer-simon-eskildsen-ceo-04u1ucc",
  "title": "Building Turbopuffer: Gergely Orosz (@pragmaticengineer ) × Simon Eskildsen (CEO)",
  "why_included": "Turbopuffer’s early architecture shows a useful agent-era discipline: ship the smallest system that tests demand, then optimize against measured hardware constraints. The transcript offers principles, not benchmarks.",
  "summary": "Turbopuffer’s first vector-search design put clustered vectors and centroids in files, fetched nearby clusters from S3, and used an Nginx cache on **one server**. Its founders drew on **eight years at Shopify**, and Cursor later adopted the database after other Postgres-based approaches.",
  "practical_implication": "Builders working with agents should keep generated architecture proportional to the current problem: start with inspectable primitives, measure latency and cost, and add bespoke machinery only after demand appears. Hardware flexibility also matters when CPU capacity is constrained.",
  "agent_context": "Turbopuffer’s first vector-search design put clustered vectors and centroids in files, fetched nearby clusters from S3, and used an Nginx cache on **one server**. Its founders drew on **eight years at Shopify**, and Cursor later adopted the database after other Postgres-based approaches.\n\nBuilders working with agents should keep generated architecture proportional to the current problem: start with inspectable primitives, measure latency and cost, and add bespoke machinery only after demand appears. Hardware flexibility also matters when CPU capacity is constrained.\n\nThe supplied transcript is incomplete and conversational. It does not provide comparative performance, cost figures, migration details, or enough evidence to generalize Turbopuffer’s architecture to other workloads.",
  "source": {
    "name": "AI Engineer",
    "url": "https://www.youtube.com/watch?v=jQDXzEVHMSE",
    "published_at": "2026-08-03T18:26:07.000Z"
  },
  "source_class": "video",
  "content_type": "Video",
  "layer": "infra",
  "domains": [
    "coding",
    "data"
  ],
  "topics": [
    "harness-engineering"
  ],
  "verification": {
    "status": "source_linked",
    "label": "Source Linked",
    "method": "source_feed",
    "verified_at": null
  },
  "uncertainty": [
    "The supplied transcript is incomplete and conversational. It does not provide comparative performance, cost figures, migration details, or enough evidence to generalize Turbopuffer’s architecture to other workloads."
  ],
  "connected_context": {
    "meaning": "This grounds the candidates’ “smallest useful system” principle at the infrastructure layer: Turbopuffer began with files, object storage, caching, and one server rather than premature distributed machinery. It adds latency, cost, and hardware availability as concrete gates for architectural escalation, while the incomplete evidence prevents treating that path as a reusable vector-search blueprint.",
    "corpus_size": 340,
    "generated_at": "2026-08-04T10:04:53.092Z",
    "connections": [
      {
        "title": "How Forward Deployed Engineering is done at Kepler — Vinoo Ganesh",
        "source_name": "AI Engineer",
        "source_url": "https://www.youtube.com/watch?v=1OMHGsUZiqA",
        "feed7_url": "https://feed7.dev/p/how-forward-deployed-engineering-is-done-at-kepler-vinoo-ganesh-0thtvuc",
        "reason": "Both favor starting with the smallest production-worthy solution and adding durable machinery only when observed demand justifies it; Turbopuffer applies that product discipline to database infrastructure."
      },
      {
        "title": "State of Data — Sean Cai, Independent / State of Data",
        "source_name": "AI Engineer",
        "source_url": "https://www.youtube.com/watch?v=ZyIoTOAbRfs",
        "feed7_url": "https://feed7.dev/p/state-of-data-sean-cai-independent-state-of-data-0v9fy69",
        "reason": "Its instruction to measure latency and cost supplies infrastructure-specific evidence for the broader candidate principle that decisions should be grounded in real operating traces rather than abstract expectations."
      }
    ]
  },
  "lifecycle": "Current",
  "published_at": "2026-08-03T18:26:07.000Z",
  "modified_at": "2026-08-03T18:26:07.000Z",
  "supersedes": [],
  "expires_at": null,
  "formats": {
    "html": "https://feed7.dev/p/building-turbopuffer-gergely-orosz-pragmaticengineer-simon-eskildsen-ceo-04u1ucc",
    "json": "https://feed7.dev/p/building-turbopuffer-gergely-orosz-pragmaticengineer-simon-eskildsen-ceo-04u1ucc.json",
    "markdown": "https://feed7.dev/p/building-turbopuffer-gergely-orosz-pragmaticengineer-simon-eskildsen-ceo-04u1ucc.md"
  }
}