The link : https://8a7ecd3c-4605-4608-a6cd-2fd5d4b739ae.weweb-preview.io/
Hey everyone, I built a real-time vocal AI agent using WeWeb and Xano, connected to OpenAI’s real-time API via WebRTC.
It supports:
Live voice input
Real-time AI response (with context)
A RAG system for dynamic knowledge injection
Still fine-tuning latency and stream handling.
If anyone has feedback on it, dont hesitate 
Wow, just tested it. Really nice work! I mean it 
What immediately caught me, each subsequent preview of voices does not stop the previous ones, so if you quickly press everything in a row, they all start screaming at the same time
+1
When I write name and click next would be nice to just click enter.
Whats the use case / purpose of this?
Thank you Batik ! Yep i thank you, gonna fix this
Fixed ! It’s basically a real-time voice assistant you can talk to directly in your browser (using WebRTC).
It uses RAG to pull accurate answers from your own docs or data so it’s not just generic GPT replies.
Could be used for support, internal tools, or even sales all in live voice, no human on the other side.
This looks sick Rayane!
When are you releasing a tutorial? 
Incredible, loved the UX, very slick. I tried with the Situational Awareness PDF and it worked very well.
The PDF is long and the model first emphasised on the fact that this essay is about cybersecurity which is only a subset of it because it’s rather about AI progress. But when I asked specific question about it, it answered very well.
Good job overall, super promising app!
Hi @rayanenocode
This is awesome. Can you show me how you have implemented this?
I am trying to build something very similar and would really appreciate if you can help.