Video unavailable. Open the source below.
NewLocal Audio Switchboard
An audio drop routes language detection, transcription, privacy and task extraction through local models.
Added to Jevfast
How it uses Jev
Inferred from the demonstration title and inspected audio interface: a single Jev call selects among local audio components such as Ear, Voz, and Redact for a dropped audio file, with about 20 decisions described. The post says there is no LLM loop, but does not expose exact request fields or question definitions.
Why it is interesting
The example suggests a typed decision layer routing an audio task through local components without a repeated general-purpose model loop. The exact routing policy remains undisclosed.