Space&Miller Voicebot
Space&Miller Voicebot is an advanced voicebot designed to demonstrate the capabilities of modern speech recognition and generation technologies. This project allows users to test the functionalities of the voicebot and see how it can be adapted to different needs.
Below is the link to the Voicebot:
Project Purpose:
The purpose of Space&Miller Voicebot is to showcase the advanced features of voicebots and enable everyone to test their capabilities. After a conversation, the bot sends an SMS to the number from which the call was made. The SMS contains a summary of the conversation and shows how the message can be tailored to the client's needs.
Key Features:
- Speech-to-Text: Utilizes Deepgram and Nova 2 technologies for accurate speech transcription.
- Speech Generation: Uses 11labs and Eleven Turbo v2 technologies to create natural-sounding voices.
- Low Latency: The bot features very low latency, at a level of 800ms.
- Personalized SMS Messages: After a conversation, the user receives a summary of the conversation in the form of an SMS, tailored to individual needs.
- Knowledge Base: The bot has its own knowledge base, created specifically for its application, making it versatile and capable of handling any situation.
- Open-ended Questions: In case of misunderstanding, the bot can ask open-ended questions to better understand the caller's needs.
- Mood Recognition: The bot is equipped with a mood recognition feature, helping it adapt to the tone of the conversation.
Technologies:
- Platform: VAPI
- Integration: Make.com
- Speech-to-Text: Deepgram, Nova 2
- Speech Generation: 11labs, Eleven Turbo v2
- Provider: Chat GPT-4 Cluster
Authors and Contributions:
- Main Author: Patryk Rogowski
- We welcome external contributions. Please report issues and submit pull requests.
Contact and Support:
For questions or issues, please contact us via GitHub Issues or directly at email:
jeremyspace@spacemillerco.com