Hey HN,
For those building apps that need meeting data (AI agents, sales tools, CRMs, etc.) – how are you handling the real-time transcription piece?
Building the necessary infra (bots for GMeet/Zoom/Teams, scaling, real-time streaming) seems like a major distraction from core product work.
We're developing Vexa as an open-source (Apache 2.0) API layer specifically for this – think of it as transcription infra you can build on.
Our v0.2 just launched a simple API for GMeet bots & real-time transcripts.
Curious about the builder perspective:
Would an open-source API like this accelerate your roadmap vs building it yourself or using closed APIs?
What are the biggest pain points in getting reliable transcription data into your application?
What features (platforms, data formats, etc.) would make such an infrastructure layer most valuable?
Looking for thoughts and experiences from fellow builders!repo: https://github.com/Vexa-ai/vexa