Google Cloud

Next 19’

live transcript

Imagine not being able to call your loved ones. Imagine not being able to call 911 during an emergency. Imagine someone was listening to every single phone call you made. You don’t have to imagine. 37.5 Million adults in the US report having some sort of hearing loss. That’s 15% of the U.S population. What do these people do? How do they place a phone call? Today, these people must use a captioning service. What that means is that if I want to call someone who is hard of hearing, I have to talk to a man in the middle, an actual human being who hears everything I say, type it, and send the text to a special landline device that the person with the hearing disability must own. Just think about it for a second, it can take almost a minute, from the moment I say something until the other side reads it. And what about privacy? This is crazy! The year is 2019. We have self-driving cars. We have rockets that can go to space and land back on earth, and still, if you are hard of hearing there’s no descent way for you to place a phone call. We want to solve that, that’s why, we created Nagish. - Nagish converts speech to text and text to speech in real time so that one side could speak and hear and the other could type and read. How we did it? I would like to call my partner, Alon Ezer to the stage to explain.

I would like to go back and talk about the captioning service Tomer just mentioned, and shared a story that happened to me. I called Jack from the New York Deaf Association to speak about a possible collaboration, tell him about our product and offer him to beta test it with us. I was connected to a captioning service that transferred everything I said to him. It started a little bit strange but very quickly became very awkward. I was telling Jack about a service that replaces the captioning service, through a captioning service! I was basically telling the translator that she might lose her job soon. I felt weird, confused, and felt I lost every aspect of privacy. I can’t even imagine how Jack feels every time he has to place a phone call. That is exactly what we want to solve in Nagish, by eliminating the middleman, we bring back privacy, independence, and transparency to phone calls, especially to those who really need it.

So now that we have Nagish and we understand the concept, how would my phone call with Jack look like today?

Well, now instead of having a real person that transfer back and forth what we both say, we use google speech-to-text API that easily converts voice into plain text. We send that text to Jack’s mobile phone and now he can read what I say.

Jack can use any text platform he wants to respond. It can be google hangouts, facebook messenger, WhatsApp, telegram and many more platforms that we will support in the future. We take the text Jack wrote and we use Google’s text-to-speech API that easily converts text into voice. We send that voice back to the phone call, so now I can hear whatever Jack type. And simple as that, we have a phone call that we both can have on the go, with no constraints and completely private. Now Jack and I are both happy.

This is pretty cool, but the cooler part is that we were able to build it in less than 12 hours. We came to a coding competition at Columbia University with an idea and using Google Cloud products, we managed to have a working demo by the end of the day.

And now I would love to show you how Nagish works.

So as you can see, we are pretty excited by this, but the question that we must is where can we go next?

We are currently working with a group of beta testers to improve our service and make it better for them. We hope to publicly launch our service by the end of the year. We are also using the Google translation API to enable our users to travel and communicate in every language they want, even if they don’t know it!

Also, we listened to our beta testers and we are going to add support for conference calls so professionals could use Nagish for their day to day work. And, we are going to add personalized voices so adults for instance, won’t have the same voice as children.

We would love to show you a live demo of Nagish so make sure you stop by the Google for Education booth right after this talk. We will be there until the end of the day.