{
  "schema_version": "1.1",
  "id": "archive:https://www.youtube.com/watch?v=XAsb7MIAzm8",
  "slug": "don-t-be-data-poor-anuj-iravane-anterior-10zyji6",
  "url": "https://feed7.dev/p/don-t-be-data-poor-anuj-iravane-anterior-10zyji6",
  "title": "Don’t be data poor — Anuj Iravane, Anterior",
  "why_included": "When production data cannot be retained, generate eval cases backward from sampled labels and reasoning paths, build records in layers, and let domain experts steer the pipeline.",
  "summary": "Anterior evaluates healthcare agents against records that are sensitive, varied, and sometimes **over 300 pages**; contracts can prohibit retaining even redacted derivatives. Its pipeline samples a label and policy reasoning path, then generates a patient journey and encounter documents from coarse to fine.",
  "practical_implication": "Reverse the inference workflow to control scenario diversity, emulate how the source documents arise, and round-trip generated records against their starting labels. Give domain experts control at each generation stage and package their reusable guidance as skills rather than leaving dataset design solely to AI engineers.",
  "agent_context": "Anterior evaluates healthcare agents against records that are sensitive, varied, and sometimes **over 300 pages**; contracts can prohibit retaining even redacted derivatives. Its pipeline samples a label and policy reasoning path, then generates a patient journey and encounter documents from coarse to fine.\n\nReverse the inference workflow to control scenario diversity, emulate how the source documents arise, and round-trip generated records against their starting labels. Give domain experts control at each generation stage and package their reusable guidance as skills rather than leaving dataset design solely to AI engineers.\n\nAnterior says **roughly 90% of its datasets** are synthetic, while clinicians distinguished synthetic from real records **about 60% of the time** in a blind review. That is a fidelity signal, not proof that the generated distribution captures production prevalence, rare failures, or clinical validity.",
  "source": {
    "name": "AI Engineer",
    "url": "https://www.youtube.com/watch?v=XAsb7MIAzm8",
    "published_at": "2026-08-19T18:00:17.000Z"
  },
  "source_class": "video",
  "content_type": "Video",
  "layer": "benchmark",
  "domains": [
    "data",
    "research"
  ],
  "topics": [
    "agent-evals",
    "skills"
  ],
  "verification": {
    "status": "source_linked",
    "label": "Source Linked",
    "method": "source_feed",
    "verified_at": null
  },
  "uncertainty": [
    "Anterior says **roughly 90% of its datasets** are synthetic, while clinicians distinguished synthetic from real records **about 60% of the time** in a blind review. That is a fidelity signal, not proof that the generated distribution captures production prevalence, rare failures, or clinical validity."
  ],
  "connected_context": {
    "meaning": "This adds a concrete way to build healthcare eval data when privacy and retention rules make real records unusable: generate documents backward from controlled labels and reasoning paths, then round-trip them for consistency. It also places dataset design with clinicians through reusable skills. The blind-review result supports surface fidelity only, leaving production coverage and clinical validity unresolved.",
    "corpus_size": 525,
    "generated_at": "2026-08-21T10:08:44.592Z",
    "connections": [
      {
        "title": "Verifiable Environments for AI in Biology — Kenny Workman, LatchBio",
        "source_name": "AI Engineer",
        "source_url": "https://www.youtube.com/watch?v=3ZMUiFaQ3qg",
        "feed7_url": "https://feed7.dev/p/verifiable-environments-for-ai-in-biology-kenny-workman-latchbio-1vs6y66",
        "reason": "Both make expert involvement necessary for domain evals; this Signal moves that involvement upstream into staged dataset generation, while LatchBio emphasizes validating graders that may reject legitimate expert paths."
      },
      {
        "title": "Don't Ship Skills Without Evals — Philipp Schmid, Google DeepMind",
        "source_name": "AI Engineer",
        "source_url": "https://www.youtube.com/watch?v=0vphxNt4wyk",
        "feed7_url": "https://feed7.dev/p/don-t-ship-skills-without-evals-philipp-schmid-google-deepmind-0fuh3ko",
        "reason": "Packaging clinician guidance as skills creates the same regression-testing obligation: their triggering and effects should be evaluated rather than trusted from manual inspection."
      },
      {
        "title": "Trading Desks to Clinical Trials: Parallels in Applied Vertical AI — Ayush Bhardwaj, Allos AI",
        "source_name": "AI Engineer",
        "source_url": "https://www.youtube.com/watch?v=Yphdry8ttAQ",
        "feed7_url": "https://feed7.dev/p/trading-desks-to-clinical-trials-parallels-in-applied-vertical-ai-ayush-1hwvsg3",
        "reason": "This supplies an implementation path for the proprietary-data constraint in vertical AI: domain experts can shape privacy-safe synthetic records when internal clinical records cannot be retained."
      },
      {
        "title": "The Regression Tax: Decomposing Why Skills Help and Hurt LLM Agents",
        "source_name": "arXiv",
        "source_url": "https://arxiv.org/abs/2607.22520v1",
        "feed7_url": "https://feed7.dev/p/2607-22520v1-0mz9wnf",
        "reason": "Expert-authored generation skills can still distort grounding or introduce regressions, so controlled labels and round-trip checks should be complemented by separate measurement of gains and harms."
      }
    ]
  },
  "lifecycle": "Current",
  "published_at": "2026-08-19T18:00:17.000Z",
  "modified_at": "2026-08-19T18:00:17.000Z",
  "supersedes": [],
  "expires_at": null,
  "formats": {
    "html": "https://feed7.dev/p/don-t-be-data-poor-anuj-iravane-anterior-10zyji6",
    "json": "https://feed7.dev/p/don-t-be-data-poor-anuj-iravane-anterior-10zyji6.json",
    "markdown": "https://feed7.dev/p/don-t-be-data-poor-anuj-iravane-anterior-10zyji6.md"
  }
}