I mean a system where one can do 10k calls per day. Has anyone built a system like this using livekit and pipecat. did you do it without using our own GPUs ?

Hi everyone, We’re exploring the idea of moving our entire voice AI stack from VAPI, Retell, ElevenLabs to LiveKit. We’re looking for a full open-source setup something more than just the server, ideally a platform that includes: - Knowledge base integration - SIP / telephony support - Live audio/video handling - Anything that makes it closer to a ready-to-use solution like VAPI Does anyone know of an open-source project that offers a full LiveKit-based platform with these features? Any recommendations, links, or experiences would be super helpful! Thanks in advance!

Hey everyone, I'm very excited to be here. I love how everyone is interested in the open source voice AI. I've been working with LiveKit for the last nine months, and I had a question. Cartesia launched their new Sonic 3 voices, however on LiveKit, when you use them, they're very flat sounding. But when you use Vapi, they have all the emotions and everything structured inside, so it feels a lot more expressive. I'm trying to understand how did Vapi achieve that? Is there a prompt in the back-end that automatically syncs up with the actual prompt people use so that the voices become more expressive? Because on LiveKit, it's actually flat.