Meet GES Delegate: Vu Van

Twitter: @vuvan

Country of Origin: Vietnam

Organization Name: ELSA Speak

Organization Website: www.elsanow.io

Brief Description of Organization:

ELSA Speak is an award-winning Artificial Intelligent mobile assistant which helps English learners speak English more confidently and correctly, using our proprietary speech recognition technology.

What inspired you to start this organization?

I was born and raised in Vietnam. 7 years ago, I moved from Vietnam to the US to pursue my MBA and Master of Education at Stanford university. Imagine coming to the US as an immigrant, and at the first job interview instead of saying “information sheet”, I said “information shit.” Imagine the influence it has on one’s career. I knew English. But speaking and being understood was an entirely different thing. I felt embarrassed and started looking into hiring a speech coach, but it was way more than I could afford. I tried to watch TV and YouTube and to mimic what I heard. But I have nobody to give feedback too.

And in this I’m not alone. There are 1.6B English learners globally who face the same challenge that I faced. From my research, I’ve also learned that people who speak English with an accent are perceived to be 30 percent less trustworthy. Company loses 30% in productivity due to mis-communication. As a result, people who speak English less confidently will earn 40 percent less income than those who do. It’s a huge problem and a huge gap to be resolved.

It’s not we don’t know english. We do! It’s just that practicing pronunciation is like exercising a muscle. Certain sounds like /th/ exists in English but don’t exist in a lot of other languages. /r/ is said differently in different parts of the world. /r/ and /l/ often interchange. So it’s not enough to know languages. You have to figure out how to say it correctly. Its not just recording yourself, its having the feedback given to you to train both your voice and your ear.

So we are building ELSA to be your private coach — it listens to you speak English, it recognizes and pick out the problem sounds that you are saying that might come across unclear to others, and it and give you the hints in a way you can learn to do and practice to improve. All done in a very simple and engaging gamified mobile interface. All of these experiences are possible thanks to our proprietary speech recognition technology that we built in-house, using machine learning and deep learning algorithm.

What is the next big step you hope to help your organization reach?

Our next big steps are to really create a fully personalized learning experience where ELSA can detect specific errors that each individual user makes, and will then send you content that focuses on those things, instead of a “one-size-fits-all” approach. We only teach the American English accent; however, we hope to expand to other accents such as British, and expand to teach other languages such as Spanish, German, French, etc.

What has been your biggest obstacle as an entrepreneur?

We’re in the process of building one of the most advanced speech technologies; however, we don’t have the same resources the Giant 5 (Apple, Google, Microsoft, and IBM Watson) do. Additionally, accuracy will always be a battle, and we think that we’ve achieved a very high rate of accuracy now, but we aren’t satisfied. We will continue to improve, especially once we expand our speech model to cater for various different accents.

What advice would you give other emerging entrepreneurs?

Get a mentor who is willing to share the entrepreneurial journey with you, it’ll change your life. A great mentor would be someone who’s willing to wake up in the middle of the night to take your call and someone who you can always lean on. My mentor was a Stanford alumni, who is now retired — he was always there to challenge me, step in to give me guidance when I ever felt confused or lost, and constantly pushed me to go beyond my limits.

And don’t be afraid to tackle a big technical challenge. As a female entrepreneur, and with a non-technical background, many people (including investors) have raised their eyebrows when I told them I am starting a tech startup in the AI / machine learning space — something that they consider is only suitable for hard-core male engineer. I believe otherwise. It’s an equal playground and as long as you have the gut to enter, you will figure out a way to make it happen. The key is to believe in yourself and don’t let others define your boundaries.