Hey everyone, just wanted to share the "why" behind this project. basically i was using twilio for a small project and realized their "intelligence" tools cost a fortune. like, $1.50 an hour just for transcripts?? that s insane for a solo dev or small startup.
So i spent the last few months building my own layer. it s a Go microservices setup that handles the audio streams and uses Qdrant for semantic search. So you can actually search for "angry customers" or "refund requests" instead of just hunting through keywords.
It ends up being like 80% cheaper than the native tools. I'm still fixing a few UI quirks but the backend is solid. Would love some feedback from anyone else who uses twilio or telnyx and is tired of the crazy bills!
Don't just record calls, understand them. AudioText Live is the missing AI layer for Twilio, Telnyx, and SignalWire. We turn raw audio streams into searchable data via real-time transcription, speaker identification, and sentiment analysis.
