Show HN: Ved AI Voice Assistant

  • Posted 2 hours ago by Amigo112
  • 1 points
https://github.com/Krish6190/Ved-Test
I built Ved because I wanted a fully offline, self hosted Jarvis for my house, that runs comfortably on everyday consumer hardware (specs 12GB VRAM GPU), without ruining system performance(multi modal).

Where I plan to take this next:

1)Server-client Architecture: I want to shift the heavy load off of ur main device onto a secondary central server where the heavy LLMS and lower models work without ruining ur personal machine.

2)Smart Home: I want to make small cheap mic/speaker endpoints (using small ups like Raspberry) allowing you to communicate with Ved without needing another appliance.

3)Remote Access: I am checking out ways to securely make Ved communicate with outside local networks for free.

I would love to hear from anyone who has worked on a multi room audio setup. What hardware or protocols you found had the lowest latency.

Please try it out and let me know if you have any feedback or suggestions :)

0 comments