_{Since Tatsu's debut on our blog in 2022 it's been quite a journey. Tatsu completed the MSc in Speech Technology and has since been hard at work with Whispp, which we'll delve into later. Along the way, he traveled to Las Vegas, appeared on NHK (Japan Broadcasting Corporation), and met the CEO of Microsoft. But let's take a step back and hear it directly from Tatsu two years after our last chat!}

Hi Tatsu! Thanks for making time for a chat with us. To kick things off, could you briefly share (once again) what drew you to study in Leeuwarden?

Sure! I come from Tokyo, Japan, where I studied a bachelor’s in Economics. I moved to the Netherlands specifically to study Speech Technology. The reason why I chose to do this study is that I was interested in learning more about speech technology and the world around it. I am very aware of the toll these disabilities take on people, so I knew I wanted to do something to help in that sense. 

The study at Campus Fryslân was really specialized and an exact match for what I was looking for. While there’s quite a few programmes out there, many of them focus on natural language processing, which is different from speech technology. However, I really wanted to focus on the algorithm behind the speech and speech AI. In this sense, this programme was quite unique in the educational landscape. 

❝I really wanted to focus on the algorithm behind the speech and speech AI.❞

Now, let’s dive into your journey with Whispp. Can you tell us a bit about Whispp and how your experience with Whispp began?

Whispp is a technology that utilizes AI to convert whispered speech and vocal cord-impaired speech into clear and natural voices in real-time. It can even recreate the patients healthy voice based on their recordings from the past.

Overall, the work done by Whispp is something I was already really interested in. My journey with them began when I simply applied via Linkedin. Once again, my motivation was to help people with speech disabilities, and who consequently cannot speak properly, to help them communicate into a natural sounding voice thanks to AI. And this is what I do at Whispp: I help develop the AI that allows people with voice-related disabilities to have a voice.

Since I joined Whispp, the team has valued the skills that I acquired through the master's programme and decided to build a close relationship with the master's programme. In fact, we have already worked with a couple of students to do a master thesis research project together.

❝My motivation was to help people with voice disabilities, and who consequently cannot speak properly, to help them communicate into a natural sounding voice thanks to AI.❞

What role do you currently hold within the company?

My official title is AI Researcher and Developer. Which means, of course, that I do research and… help develop! To put it simply, I work with problems, and think about how to solve them. To do that, I review the research, narrow down the approach, and ultimately develop the model. I then test it to see if we can use it for the product. This process goes on endlessly.

At the moment, we provide the app directly to consumers, that is people with disabilities who need help. Just at the start of 2024, we launched a phone call functionality and we are currently exploring partnerships. Last month we went to CES (Consumer Technology Association), in Las Vegas. There, each company has a booth to show their technology. We won an award for that and got a lot of media attention from all over the world. One of them was from the Japanese national television!

Reflecting on your time in our Master’s programme, what specific skills or knowledge do you feel it equipped you with that have been valuable in your current position at Whispp?

Because I came from Economics, I have to say, it was quite tough. Studying took most of my time, but it’s definitely not impossible! Because it is so specialized, you can position yourself well in the market, so if you are really specialized in this field, it does pay off.

We are talking about a very specialized field, even if you are familiar with AI. Because of the lack of knowledge in the audio field, sometimes they find it difficult to get in. In Speech Technology, I learned the fundamentals, i.e. what audio is. e. The fundamentals are key for you to thrive in this field. Without them, you cannot move up to more advanced knowledge.

Speech: how do you decompose the speech? Sounds have a lot of properties; how do humans actually generate a voice? There’s a model we follow to generate speech. This is not just knowledge, there’s theories behind it. It is not really related to AI, but the truth is, you need to know that to be able to work with AI. 

What advice would you offer to someone who's uncertain about applying to this programme? What do you think makes it worth choosing?

It really depends on the person’s motivation. If they are really interested in speech tech, then they should go for it. But there’s a trade off. Because it is so specialized, it might be difficult to step away from this field and do something different. So if you are really into this field, I think this master’s provides you with a way in! However, if there is any doubt, then you really have to consider.