Microsoft, MAC, Lewis discussed on The Voicebot Podcast


Absolutely love this conversation. This idea particularly with how important latency is in voice in really just saving a couple hundred milliseconds in a voice conversation. And it's actually going to be more than that if you're if you're going round trip to another To another cloud service provider. But that can make a real difference in in that particularly also thinking about other services are other features you might implement if for example Bargin so if you're going to allow people to interrupt the assistant is is lazy is really important without. It's it's always important but it's really important with that So so i really liked that conversation So when we think about the teneo tech stack and microsoft so lewis is obviously very important part of that As your other services using from the microsoft stack the it listeners might be interested in voice absolutely and then we're also try links some of the There's face recognition that you can get there now and also emotional detection but those are those are still quite early so the face recognition. Worst quite. well if you have a. I can't remember what it's called. But it's a microsoft device which took sorta like an mac and it's quite high-powered if you have that with a really good camera on it. It works but i wouldn't be able to do it on my camera. It's difficult to do it on across many devices. But that's definitely coming definitely comey where you will have face facial recognition to get emotions and know when to connect this person into a human. Instead of having talked to a bolt. i would say be. I don't know if you try to solution but it's also things like the might will get things back from the api. Which are things like man in his fifties who seems to be happy. I these these type of things that they detect and there's going to be more useful. I think yeah so are you. Let me ask you a different. Let me redirect question. Are you hearing from your customers in interest in having the emotional recognition as well as intent recognition. Very much very much. We still do it based on text right. So there's this analogy piece. We have a really found a good way to do it onto novelty. There's some way to do that. In retrospect but not a real time and which sort of defeats the purpose. You wanna do it in real time. But that's probably coming as where you'll be able to judge the tenacity and the actual words shoes but now it's based on words right..

Coming up next