The Voximal stack

The Voximal Stack: Open-Source Voice Assistant Platform

An open-source alternative to proprietary voice assistant platforms like Alexa or Siri, allowing developers to build custom voice assistants with speech recognition, NLP, and TTS technologies.

What is The Voximal stack?

The Voximal stack is an open-source platform for building voice assistants and conversational AI applications. It provides developers with speech recognition, natural language understanding, dialogue management, and text-to-speech capabilities to power the next generation of voice interfaces.

At the core of Voximal is Rhasspy, an offline, privacy-first voice assistant toolkit. Rhasspy handles wake word detection, automatic speech recognition (ASR), intent classification, entity extraction, dialogue management, and text-to-speech synthesis. It supports languages like English, French, German, Spanish and more.

Buildling on top of Rhasspy, Voximal adds client-server architecture, device integration, multi-user support, skills framework, and tools for authoring conversational content. The modular design allows customization of each component like ASR, NLU, or TTS.

Key capabilities offered by Voximal:

• Custom wake word training
• Speaker identification and verification
• Dynamic speech recognition
• Advanced natural language capabilities
• Flexible dialogue management
• Multilingual TTS voices
• Skills SDK for integrating third-party services
• Tools for authoring rich conversational content
• Multi-device orchestration over LAN or Internet
• User and device management interfaces
• Open API for interoperability with other systems

With Voximal's privacy-focused approach, all user voice data stays on local devices. The platform is fully open-source under GNU affero GPL 3.0 license and backed by an active open community.

The Voximal stack Features

Official Links