Inspiration
Just a fun project to understand what mighty birds may think and say to you.
What it does
Interacting with a voice agent to hold a conversation and retrieve needed information or have a fun conversation.
How we built it
Mainly the credits from Rime.ai were used to build an interactive profile. The Python connects the API and the interactive web page, aka chat, for the speech. The transcription is included for in-and-out conversation. For the demo overview some "vibes" were included. Check them out. VAPI voice packages were explored to give a concept; Mr. Goose is speaking in option 1: Bubba Marshal, and option 2: Road Dawg voice (other options tested were the Giovanni voice, etc.). Please rate what voice from VAPI you want to hear from Goose the most.
Challenges we ran into
The main points to overcome and improve in the future are the speed of the voice and reduction of the wait time in between "thinking" (token generation) time and the actual response.
Accomplishments that we're proud of
The demo can be seen in the Figma pages. The understanding of the concepts for speech recognition and processing is the main accomplishment of our work.
What we learned
Voice activity detection, speech-to-text, and text-to-speech.
What's next for Voice Goose
Broaden space; make it even more interactive through the video support. Adding the graphic for a digital goose and creating a flow in speech.
Built With
- amazon-web-services
- figma
- huggingface
- python
- remix
- rime.ai

Log in or sign up for Devpost to join the conversation.