One book, many voices
The book's owner turns on collaboration and invites narrators by email. An invited narrator with a plan is added automatically, and the shared book simply appears in their Start screen. Everyone sees the same manuscript, the same cast, and the same chapter rail.
You can only record your own lines
Casting decides everything. Assign characters to narrators, and each person's assigned lines light up for them; everyone else's lines are read-only. That rule is enforced by the server, not just the interface, so takes can never collide: your line, your take, your name on it. Lines carry each narrator's avatar, presence shows who is in the chapter right now, and a recording ring marks whoever is at the mic.
Playback and export understand the cast
Play a chapter and Medrak interleaves everyone's takes in script order, with silent gaps standing in for lines not yet recorded. Export the finished book as one blended mix, as separate per-narrator stems, or as by-narrator run chunks that drop straight into an editor's session. Cloud copies of every take are stored lossless, so what the editor downloads is bit-for-bit what was recorded.
Where it fits
Collaboration works in the browser and the desktop app. The Narrator plan collaborates on shared books; the Studio plan runs full productions with per-seat billing for the whole cast. Every 14-day trial includes it, so a duet is the fastest way to feel it. The audiobook workspace post covers the prep and casting that comes first.