# hearing

> Pluggable meeting transcription & context-aware AI agents, batch or live


## Pages in this subsection

- [hearing](hearing.html.md): hearing — pluggable meeting transcription & context-aware AI agents.
- [hearing.agents](hearing.agents.html.md): The agent / “copilot” layer — the actual point of the hearing project.
- [hearing.backends](hearing.backends.html.md): Extra scribed backends that hearing contributes: Apple-Silicon MLX and WhisperX.
- [hearing.backends.mlx_whisper](hearing.backends.mlx_whisper.html.md): mlx-whisper (Whisper on Apple Silicon via MLX) backend for scribed, from hearing.
- [hearing.backends.mlx_whisper.adapter](hearing.backends.mlx_whisper.adapter.html.md): Adapter for the mlx-whisper backend (Whisper on Apple Silicon via MLX).
- [hearing.backends.mlx_whisper.config](hearing.backends.mlx_whisper.config.html.md): Configuration for the mlx-whisper backend (Whisper on Apple Silicon via MLX).
- [hearing.backends.parakeet_mlx](hearing.backends.parakeet_mlx.html.md): parakeet-mlx (NVIDIA Parakeet-TDT on Apple Silicon via MLX) backend for scribed, from hearing.
- [hearing.backends.parakeet_mlx.adapter](hearing.backends.parakeet_mlx.adapter.html.md): Adapter for the parakeet-mlx backend (NVIDIA Parakeet-TDT on Apple Silicon).
- [hearing.backends.parakeet_mlx.config](hearing.backends.parakeet_mlx.config.html.md): Configuration for the parakeet-mlx backend (NVIDIA Parakeet-TDT on Apple Silicon).
- [hearing.backends.whisperx](hearing.backends.whisperx.html.md): WhisperX (batched Whisper + alignment + optional diarization) backend for scribed, from hearing.
- [hearing.backends.whisperx.adapter](hearing.backends.whisperx.adapter.html.md): Adapter for the WhisperX backend (transcribe -> align -> optionally diarize).
- [hearing.backends.whisperx.config](hearing.backends.whisperx.config.html.md): Configuration for the WhisperX backend.
- [hearing.capture](hearing.capture.html.md): Audio capture and channel handling for the hearing pipeline.
- [hearing.cli](hearing.cli.html.md): Command-line interface for hearing, dispatched with `cw`.
- [hearing.context](hearing.context.html.md): Context stores and retrieval (RAG) for context-connected agents.
- [hearing.diarize](hearing.diarize.html.md): Diarization and speaker labelling for the hearing pipeline.
- [hearing.interfaces](hearing.interfaces.html.md): Facade Protocols for the four pipeline concerns.
- [hearing.pipeline](hearing.pipeline.html.md): Pipeline assembly — compose the four concerns by dependency injection.
- [hearing.storage](hearing.storage.html.md): Transcript persistence via key-value stores (python-storage / `dol`).
- [hearing.stt](hearing.stt.html.md): Speech-to-text for hearing — now backed by the scribed transcription façade.
- [hearing.types](hearing.types.html.md): Shared data model for the hearing pipeline.
- [hearing.vad](hearing.vad.html.md): Voice-activity detection and utterance segmentation — re-exported from scribed.

## Optional

- [Top-level llms.txt](../llms.txt): Complete documentation index.