We worked on building a real-time voice to voice desktop translation app that worked with all desktop video calling apps including Zoom, Meet and teams. The development required figuring out desktop level plumbing with virtual drivers like blackhole & vb-cable for voice routing. We also set up up and optimised the latency for voice translation pipeline. After working on it for 6 months, Google announced launch of native voice to voice translation capabilities in their last IO. Zoom and Teams followed the suite. We stopped actively working on it post that but we still get dozens on demo requests on the website - hence decided to opensource the project - in the hope that this might be taken up by the community
We worked on building a real-time voice to voice desktop translation app that worked with all desktop video calling apps including Zoom, Meet and teams. The development required figuring out desktop level plumbing with virtual drivers like blackhole & vb-cable for voice routing. We also set up up and optimised the latency for voice translation pipeline. After working on it for 6 months, Google announced launch of native voice to voice translation capabilities in their last IO. Zoom and Teams followed the suite. We stopped actively working on it post that but we still get dozens on demo requests on the website - hence decided to opensource the project - in the hope that this might be taken up by the community
We also launched on ProductHunt and were rated one of the highest rated products of the day: https://www.producthunt.com/products/ztalk-ai
Here's a demo of the product: https://www.youtube.com/watch?v=tIZ1kDEMjP8