Just dropped my first YouTube video showing Sero's Call Assist module in action.
What it does:
Takes audio from any call, transcribes it live using Whisper, detects when someone asks a question, and suggests a personalized response in about 2 seconds.
This was built in a single late night session using the vibe coding method. No traditional programming. Just orchestrating AI tools to build something that actually works.
The video is short, less than 2.5 minutes. You can see the question get detected and watch the response appear almost instantly.
This is the kind of stuff we are building here. Real tools that solve real problems. Not tutorials that sit in a folder. Working systems.
If you have questions about how it works or want to see the full build process with all the debugging and problem solving, let me know. I recorded the whole 45 minute session too.
More coming soon.