startDictationSession
Open a live dictation session: a private room the caller's microphone streams to, which sends back the words as they are spoken. Each recording takes a session of its own, and every call makes a new room, so the answer must never be cached. NOT_CONFIGURED with service: VOICE where live dictation is not set up, and the app then uploads the recording instead (POST /api/voice/transcribe).
startDictationSession: StartDictationSessionPayload!Type
StartDictationSessionPayload object
A live dictation session, as @recruiteros/voice reads it.
startCallRoleCheck
Check a call against the workspace's open jobs. The work is done in the background, after this request has been saved: the answer comes back with the check PENDING, and `Call.roleCheck` is polled every couple of seconds until it is COMPLETED or FAILED. Without `force`, a finished check is answered as it is and nothing is started. Asking again while a check runs starts nothing more. The check runs on the model, so what it spends is kept whatever becomes of later requests.
startPursuit
Start pursuing a company through a person there, from their LinkedIn profile: the company becomes a lead if it is nothing yet (a client stays a client), the person becomes its contact, and the pursuit is recorded, owned by the caller and at `PURSUE`. The contact is the workspace's contact with that profile, moved to this company if they were elsewhere and given the name and role sent. Asking again for the same person at the same company returns the pursuit already there, adding the advert if it is new. Needs a user to act as.