Hacker News new | ask | show | jobs
by JustinGu 851 days ago
Wow this is incredible. I've worked a bit in the conversational LLM space and one of the hardest problems we were struggling with was human interruption handling. From the demo it seems like you guys have it down. Can't wait to see where this goes :) BTW I don't think the demo works on mobile, tried it on safari on IOS and got no response.
2 comments

They're most likely have two "agents" working in tandem to listen and speak, and it seems like the listener takes precedence over the speaker agent but underneath they share the same context window. Programming wise, probably using multithreading and channels architecture, depending on the programming language.
What a cool trick to pull that off!
It might ask for permission to use the microphone. If you can't find it, try going to the website's homepage, where you can enter your phone number to receive a call.
https://github.com/vocodedev/vocode-python

If your looking for phone flavour, also very good. Curious which is 'better'

Retell is much stronger at handling human interruptions
Yep, I gave permission for both my mac and phone but I got to try the demo out on mac anyways
It seems to be broken on iOS Safari. I got no response after accepting the microphone prompt.
Thanks for the feedback. We will look into it