Mnatsakan Sharafyan (MSE ‘18): Introducing the World’s First AI-based Personalized Adaptive Speed Solution

5 min read

Mnatsakan Sharafyan (MSE ’18) is a graduate of the American University of Armenia (AUA) Master of Science in Economics (MSE) program. After mastering mathematical and statistical thinking, Mnatsakan became interested in data science, which led him to pursue a second master’s in applied science at the University of Michigan. Mnatsakan is the co-founder and COO of Saima, the world’s first AI-based personalized adaptive speed solution startup, which aims to revolutionize online learning and video watching. Through the use of neuroscience and data science, Saima automatically adapts the speed of the video one is watching to one’s cognitive capabilities and trains the brain to reach even better results in less time. Saima’s recently launched Chrome extension can be used for online learning videos on popular platforms like YouTube, Coursera, edX, and Loom. 

In the interview below, Mnatsakan talked to us about his career path following graduation from AUA, the struggles he experienced studying online, and how the startup idea of Saima was born. 

What inspired you to continue your education at AUA?

After graduating from the Armenian State University of Economics with a bachelor’s degree in business, majoring in accounting and auditing, I understood that there are more questions that still needed to be answered, as education for me is the continuous process of finding answers to your questions and defining new ones. At that time, I knew I had to continue my studies to gain more profound knowledge in micro and macroeconomics. That is when I started researching options for a master’s degree in economics. In that process, I attended open-door events at AUA, during which I heard several presentations and learnt that the MSE program is one of the most demanding master’s programs at AUA, which made me want to rise to the challenge.

What was your favorite class, and who was your favorite professor?

Dr. Gayane Barsegyan’s Macroeconomics course was my favorite. I loved it because there was a lot of math and struggle, which is what I enjoyed the most. Another subject I enjoyed was asset allocation and multi­ asset investing. I consider myself one of the luckiest students that had a chance to listen to lectures by Joseph Simonian, instructor of that course then. His thoughts and insights on financial markets were invaluable. Simonian is currently an official advisor to Saima.

What skills have you learned at AUA that have served you well throughout your career? Where did life take you after you graduated from AUA?

For me, the most important skills I gained at AUA were from the quantitative courses. If you have an inclination for mathematics and statistics, you can quickly adapt to other disciplines and easily work your way through the coursework. That’s why I value the mathematical and analytical skills I acquired at AUA. The second most valuable thing that AUA gave me is the network of contacts. There, I made many good friends I can rely on to this day. Many of the friendships forged at AUA later turned into partnerships.

After graduation, I worked as a business risk operations manager at HSBC Bank Armenia. Then, I moved to the public sector and worked as head of the Customs Statistics and Income Registration Department at the State Revenue Committee of the Republic of Armenia. And finally, all those career changes brought me to where I am now: working on my startup. Academically, I also pursued several certificates in risk management and then, in 2020, I enrolled in the master’s program at the University of Michigan School of Information, majoring in applied data science. This degree allowed me to pivot from economics to data science.

Tell us about Saima. How did you come up with the idea for the startup?

The story behind Saima is a personal one. It all started with my struggle when studying at University of Michigan. I started school right when Covid-19 hit the world, so all my classes were held online. Soon, I faced a problem: I needed help to keep up with the pace of the video lectures. As a non-native English-speaker with some attention issues, I struggled with online classes, particularly with watching pre-recorded video lectures, which nowadays has the major stake in online learning. That’s when I started to think about how to solve the problem I was facing and increase the efficiency of video consumption — understanding better in less time.

I knew I was not alone and that millions of non-native English speakers would have similar struggles. Statistics show that 1.5 billion people speak English, of whom only 400 million are native speakers, meaning 1.1 billion people in the world speak English as their second or third language. I knew this was a global issue. Thus, my co-founder Shushan Nersisyan and I started to research the field, finding that many learners face challenges because of the video/audio speed. We focused on that issue and started to look for ways to address it. That’s how Saima was born. 

What is the difference between simple playback speed functionalities and Saima?

Playback speed functionalities are just a brute force that multiplies and divides the speed by one factor, which allows you to divide or multiply the rate of speech of the speaker — a speed which most often is not conducive to good comprehension, memorization, or focus because we all speak at different rates even in the same video. For example, in Coursera or Youtube tutorial videos with a single speaker, the speed of speech fluctuates from 110 to 220 words per minute. Hence, if you divide or multiply by a single factor over the entire video, you would not have the most efficient speed for your comprehension, focus, and memorization. At Saima, we ensure the right speed for every point of video. So the speech changes every fifty milliseconds to match your comprehension, which we derive from the cognitive questionnaire user answers in advance. Thus, as you watch a video, Saima automatically speeds up and slows down to match your cognitive capabilities.

What is the primary goal of your product? And who are the beneficiaries? What challenges did you face when creating Saima?

We aim to increase video and audio comprehension, memorization, and focus, which can be a real pain for non-native English-speakers. Our beneficiaries are from all over the world, but we are currently working mainly with India, Brazil, and Mexico. Also, students with attention deficit hyperactivity disorder can benefit significantly from using Saima.

It happened that we started this venture in the pre-recession market, which subsequently slowed down the process. We had some challenging times, starting with forming the team and adapting our business cycles to the new reality. We were able to raise the money needed for the business.

What is your definition of success? What achievement are you most proud of?

Success is the achievement of accurately pre-defined goals. Why accurately? Because you can define some goals for yourself, achieve them, and then understand that you had defined them wrongly. So, correctly defined goals are crucial for real success. For me, personal success is waking up and thinking about all the people whose life I am positively impacting. Thus, Saima is the achievement I am most proud of.

Are you thinking about new startups/products in the future?

I love the process of inventing, innovating, and building things, so there will definitely be new startups in the future.

What would be your advice to people who want to build a startup?

For me, the main guarantee for a successful venture is the team. A good team is the number one guarantee for success. I am lucky to have my colleagues and friends at Saima, who are great professionals in their fields. I am happy we have teamed up together to create Saima.