How your briefing episodes are generated

When you press Create podcast, Quorum produces one personalised episode for every attendee of the meeting. Understanding the stages helps you read the progress on the meeting page and know what to expect.

The stages

1. Reading the files. Quorum fetches every file you assigned and extracts its text. Files that cannot be read as text, such as a scanned image-only PDF, are skipped with a note in the logs, and the rest carry the briefing.
2. One shared summary. The selected documents are condensed once for the whole meeting. This pass takes special care with notes and minutes of previous meetings, preserving decisions, action items with their owners, and open questions, and preferring newer documents when sources conflict.
3. Planning all episodes together. Before any script is written, an editorial planner coordinates every attendee's episode so no two listeners are told conflicting things. Facts belong to everyone. Only the recommendations, emphasis, and depth are tailored per listener.
4. Writing and voicing, one attendee at a time. Each attendee gets a script tailored to their saved preferences and their slice of the plan, then the script is rendered to audio with distinct host voices. Episodes render sequentially, so with many attendees the later episodes take longer to appear.

Following progress

The meeting page updates live. While generation runs you will see a Rendering card with each attendee's episode and its status badge: Queued, Generating, Ready, or Failed. Each episode becomes playable the moment it is ready, so early attendees can start listening while later episodes are still rendering.

The meeting finishes as Completed when at least one episode succeeded. If a single episode fails, the others are unaffected and the failed one shows its own error message.

The ready email

When an attendee's episode completes, Quorum emails them a link to listen. Emails only go to attendees who have a Quorum account. People outside the workspace are never emailed.

How episodes are personalised

Each script is shaped by three layers working together:

• Your workspace's Podcast settings set the show identity, style, formality, humour, depth, language, and content rules.
• The listener's own Preferences (domain, role, interests, and custom instructions) steer what their episode emphasises.
• The meeting's goals focus every episode on what the meeting needs to achieve.

The hosts are outside voices briefing the listener. They speak about the meeting and the team in the third person, address the listener directly, and never take on work themselves.

Length and timing

There is no target duration. An episode's length follows the content: the hosts cover what materially helps the listener prepare and then stop. Generation time depends on the number of files, the number of attendees, and audio rendering, so a meeting with many attendees can take several minutes to complete end to end.