Credits & third-party licences

Voice. Maya speaks with the Piper voice en_GB-aru-medium, trained on the ARU speech corpus (Anglia Ruskin University / University of Liverpool), distributed under the Creative Commons Attribution 4.0 International (CC BY 4.0) licence via the rhasspy/piper-voices collection. The voice is a generic synthetic speaker; no real person is imitated.

Speech synthesis: Piper (piper-tts, GPL-3.0; run server-side only, unmodified).
Speech recognition: faster-whisper (MIT) running OpenAI Whisper base.en weights (MIT) via CTranslate2 (MIT).
Voice activity detection: Silero VAD (MIT), ONNX Runtime (MIT).
Web: FastAPI, Starlette, Uvicorn (MIT); httpx (BSD-3); qrcode (BSD); Pillow (MIT-CMU); NumPy (BSD-3); PyAV (BSD-3, bundles FFmpeg LGPL/GPL components).
Infrastructure: Caddy (Apache-2.0), Python (PSF), Docker images as tagged.

The full list with licence texts is in THIRD_PARTY_NOTICES.md in the source distribution.