Aivah
Talk to sales

Productivity

Podcasts

Generate two-voice podcasts from an agent's knowledge

Turn an agent's knowledge into a two-voice podcast: a host that interviews and steers, and an expert that answers. The resulting audio plays back inside Aivah and is available globally through AI Drive.

Open from Side nav → Productivity (/productivity) and pick Podcast in the Type selector.

Configure

Click Edit in the productivity toolbar to open the podcast dialog.

SettingOptionsWhat it controls
LengthShort (~6–10 min), Default (~15–22 min), Longer (~26–40 min)Approximate duration. Short for a snappy intro, Longer for an in-depth conversation.
Host voiceAny voice from the libraryThe voice of the curious co-host who interviews.
Expert voiceAny voice from the libraryThe voice of the clear-explainer subject-matter expert.
Focus (optional)Free textGuide what the hosts should emphasize (e.g. Focus on chapter 3 and practical examples). Leave blank to cover the full source material.

Voice options include Aivah's built-in voices as well as any voice clones you've added.

Podcast edit settings
Edit Podcast Settings – Length, Host / Expert voices, and an optional Focus prompt

Workflow

  1. Pick the source agent

    Use the Agent selector in the toolbar. The infinite-scroll search filters quickly across large libraries.

  2. Configure length and voices

    Click Edit, choose a Length, then pair a Host voice with an Expert voice.

  3. Generate

    Click Generate. While the episode is rendering, the Document Creating Banner appears in active conversations for that agent.

  4. Play it back

    Once ready, the podcast becomes accessible from the floating Podcast player in Avatar chat for the same agent, and from the AI Drive gallery.

Where podcasts live

  • Avatar chat → Podcast player (floating toolbar) – the fastest way to listen in-context.
  • AI Drive – global browse and download.
Podcast player join
Podcast player – press Join to ask a question and have the hosts respond live
Podcast player listening
Podcast player listening – it captures your question and waits for you to pause

Voice-pairing tips

  • Use distinct voice families – pair a calmer host voice with a more energetic expert voice (or vice versa) to make turn-taking obvious.
  • Match accents to your audience – a regional voice can make technical content feel more accessible.
  • Test with short first – generate a short episode to validate the voice mix before committing to a long run.

Related guides