{
  "schema_version": "1.1",
  "id": "s8:https://www.youtube.com/watch?v=V-isu4eTHgw",
  "slug": "connect-ai-to-billions-of-legal-documents-simon-eskildsen-turbopuffer-ja-1xhg8ow",
  "url": "https://feed7.dev/p/connect-ai-to-billions-of-legal-documents-simon-eskildsen-turbopuffer-ja-1xhg8ow",
  "title": "Connect AI to Billions of Legal Documents — Simon Eskildsen, turbopuffer & Jacob Lauritzen, Legora",
  "why_included": "Legora’s latency failure shows why tenant activity patterns matter more than schema shape: mixing hot and cold projects caused cache thrashing, while per-project namespaces restored isolation.",
  "summary": "Legora’s partitioned Postgres design mixed active and dormant projects, causing cache thrashing as the corpus grew. Search and ingestion P99 rose from **100 milliseconds to 20 seconds** before the company moved at roughly **400 million documents**.",
  "practical_implication": "For multi-tenant agent retrieval, align the storage boundary with the unit users query. Legora adopted one Turbopuffer namespace per project, gaining BM25 search, per-namespace separation, and object-storage economics for cold projects across **more than 2 billion documents**.",
  "agent_context": "Legora’s partitioned Postgres design mixed active and dormant projects, causing cache thrashing as the corpus grew. Search and ingestion P99 rose from **100 milliseconds to 20 seconds** before the company moved at roughly **400 million documents**.\n\nFor multi-tenant agent retrieval, align the storage boundary with the unit users query. Legora adopted one Turbopuffer namespace per project, gaining BM25 search, per-namespace separation, and object-storage economics for cold projects across **more than 2 billion documents**.\n\nThe design fits workloads with many cold indices and a small hot set; it is not a universal database prescription. Cold reads can still incur object-storage latency, and some Legora workloads disabled the NVMe cache to meet isolation requirements.",
  "source": {
    "name": "AI Engineer",
    "url": "https://www.youtube.com/watch?v=V-isu4eTHgw",
    "published_at": "2026-09-16T16:00:19.000Z"
  },
  "source_class": "video",
  "content_type": "Video",
  "layer": "infra",
  "domains": [
    "data"
  ],
  "topics": [
    "retrieval",
    "enterprise"
  ],
  "verification": {
    "status": "source_linked",
    "label": "Source Linked",
    "method": "source_feed",
    "verified_at": null
  },
  "uncertainty": [
    "The design fits workloads with many cold indices and a small hot set; it is not a universal database prescription. Cold reads can still incur object-storage latency, and some Legora workloads disabled the NVMe cache to meet isolation requirements."
  ],
  "connected_context": {
    "meaning": "This adds a concrete storage lesson to enterprise retrieval: partitioning must follow the user-visible query and isolation boundary, especially when a small hot set sits over many cold corpora. The migration supports separating metadata and retrieval workloads, but its object-storage tradeoffs narrow the design to similarly skewed multi-tenant systems rather than a general database replacement.",
    "corpus_size": 812,
    "generated_at": "2026-09-19T09:06:04.843Z",
    "connections": [
      {
        "title": "How Web Data Infrastructure Powers the Next Generation of AI — Patricija Žemaitytė, Oxylabs",
        "source_name": "AI Engineer",
        "source_url": "https://www.youtube.com/watch?v=1UmZHb_E_SM",
        "feed7_url": "https://feed7.dev/p/how-web-data-infrastructure-powers-the-next-generation-of-ai-patricija-z-1ta4xzh",
        "reason": "Oxylabs establishes retrieval as production-critical infrastructure; Legora identifies corpus layout and cache behavior as another source of latency collapse beyond browsers, proxies, parsing, and retries."
      },
      {
        "title": "Serving 2 Million Models Without Melting: Scaling the Hugging Face Hub — Arek Borucki, Hugging Face",
        "source_name": "AI Engineer",
        "source_url": "https://www.youtube.com/watch?v=lyL5QhgIOxc",
        "feed7_url": "https://feed7.dev/p/serving-2-million-models-without-melting-scaling-the-hugging-face-hub-ar-06hazr6",
        "reason": "Both architectures separate workload paths instead of scaling one undifferentiated store; Legora applies that principle to per-project indices and hot-versus-cold document access."
      },
      {
        "title": "Why Your Enterprise Tech Stack Isn’t Ready for AI Agents — Christopher Lovejoy & Saul Howard",
        "source_name": "AI Engineer",
        "source_url": "https://www.youtube.com/watch?v=mav15aW9lLM",
        "feed7_url": "https://feed7.dev/p/why-your-enterprise-tech-stack-isn-t-ready-for-ai-agents-christopher-lov-1uwgsr1",
        "reason": "Legora provides an implementation consequence of treating enterprise constraints as foundational: tenant isolation can determine namespace design and may require disabling a performance cache."
      },
      {
        "title": "Regional inference now available on AI Gateway",
        "source_name": "Vercel",
        "source_url": "https://vercel.com/changelog/regional-inference-now-available-on-ai-gateway",
        "feed7_url": "https://feed7.dev/p/regional-inference-now-available-on-ai-gateway-0642f58",
        "reason": "Regional routing and per-project namespaces enforce different boundaries—residency versus retrieval isolation—but together show that sensitive workloads constrain infrastructure placement before model execution."
      }
    ]
  },
  "lifecycle": "Current",
  "published_at": "2026-09-16T16:00:19.000Z",
  "modified_at": "2026-09-16T16:00:19.000Z",
  "supersedes": [],
  "expires_at": null,
  "formats": {
    "html": "https://feed7.dev/p/connect-ai-to-billions-of-legal-documents-simon-eskildsen-turbopuffer-ja-1xhg8ow",
    "json": "https://feed7.dev/p/connect-ai-to-billions-of-legal-documents-simon-eskildsen-turbopuffer-ja-1xhg8ow.json",
    "markdown": "https://feed7.dev/p/connect-ai-to-billions-of-legal-documents-simon-eskildsen-turbopuffer-ja-1xhg8ow.md"
  }
}