{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreibbaoargkic4wqz2v7wo3zj2edec2zyqijyirylohmien7clfzpku",
    "uri": "at://did:plc:lk3jfj3zq4k4wxnk474axylu/app.bsky.feed.post/3mo4ohwsm6at2"
  },
  "path": "/t/multi-agent-voice-mode-for-thesis-defense-simulation-and-research-training/1383361#post_2",
  "publishedAt": "2026-06-12T20:56:09.000Z",
  "site": "https://community.openai.com",
  "tags": [
    "@eng-sajad."
  ],
  "textContent": "Thanks for expanding on the idea, @eng-sajad. The thesis defense example makes the use case much clearer.\n\nWhat you're describing goes beyond simply switching voices. It would allow multiple AI participants in the same live Voice Mode session, each with their own expertise, personality, and questioning style. That could be especially useful for thesis defenses, oral exams, technical interviews, engineering training, and other situations where realistic panel-style discussions matter.\n\nI've passed this feedback along for logging so the team can review it. The detailed academic example and clear distinction from the current single-speaker experience are exactly the kind of context that helps illustrate the value of a feature request.\n\n-Mark G.",
  "title": "Multi-Agent Voice Mode for Thesis Defense Simulation and Research Training"
}