Hi everyone! I'm new to managing tools for our distributed team and I'm a bit out of my depth here.
We have a 200-person company with teams across four countries. Our biggest pain point is real-time translation during video calls—both for spoken word and live chat. We've tried a couple of basic AI translation APIs, but they either lag too much for a live conversation or get really expensive at our scale.
Could anyone share what's actually working in production for this? I'm especially curious about:
- Which LLM providers handle this streaming/real-time aspect best?
- How do you balance cost against latency when you have so many concurrent users?
- Is the quality good enough for fast-paced technical discussions, or does it still stumble?
Any real-world examples or gotchas would be super helpful! 😅