{
  "id": 11531735,
  "title": "Moderating Spoken Content Explained (Why Support Teams Transcribe Text First)",
  "url": "https://urgent.news/2026/10/02/moderating-spoken-content-explained-why-support-teams-transcribe-text",
  "topic": "tech",
  "section": "Tech",
  "published": "2026-10-02T21:39:58.000Z",
  "source": {
    "name": "Dev.to",
    "slug": "dev-to",
    "url": "https://dev.to/frosty45/moderating-spoken-content-explained-why-support-teams-transcribe-text-first-3p5b"
  },
  "original_language": "en",
  "account": null,
  "summary": "The article discusses the importance of transcribing spoken support requests before moderating the text. It emphasizes that a direct audio-to-label shortcut may be faster, but it lacks the necessary auditability and reproducibility required for a marketplace ticketing system. The author suggests making transcription a durable stage, consuming an immutable transcript, and ensuring both stages are idempotent. The article also highlights the need to distinguish between transcription and moderation results, as they answer different questions. A structured output contract is proposed for the ticketing application, containing a schema version, a closed label set, evidence spans from the transcript, and an action. This approach ensures that only valid and reliable decisions are made, and any failure states are clearly communicated.",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}