Advice from International Alumni: Muhammad Sarmad

The KAIST Herald interviewed KAIST international alumni about their experiences as students in KAIST and how these experiences have helped them in their current careers. This interview series was conducted in collaboration with International Scholar and Student Services (ISSS). 



Please introduce yourself.

My name is Sarmad, and I did my masters in electrical engineering at KAIST. I started in 2017 and I was in Professor Hyunjoo Jenny Lee’s lab.


Why did you specifically choose Electrical Engineering in KAIST?

Before my masters, I did my bachelors in electrical engineering, and my major was in control system design. Afterwards, I was hoping for some sort of change. I worked in the industry for about two or three years in Pakistan, and I wasn’t exactly sure what I wanted to study but I was sure that I didn’t want to study control systems the way I was doing that time. Therefore, I applied to a bunch of programs, and at that time, KAIST’s program seemed general enough that you can just apply and see what happens. You do actually have to choose a professor at the time of application, but you can basically just see when you get accepted. So that’s why I went with this program, and when I came to KAIST, I took a whole bunch of courses. I not only took courses related to biomedical microsystems, but I also took courses related to artificial intelligence, and I soon found that my interests lie in the artificial intelligence, deep learning, and computer vision research areas. Then from there onwards, I started going towards that path and pursued my thesis also in that direction.


Along that topic, you said that you took some classes that made you understand you wanted to pursue a field in artificial intelligence. What specific activities or classes did you take that made you more sure that you wanted that field?

When I came to KAIST, I didn’t have a clue what artificial intelligence was and what it could mean for you, as well as what Korean artificial intelligence looks like. So I started with the very basic courses, and I remember the first ever course I took was a very basic course about deep learning and AlphaGo. Just imagine, I was [in my] first semester in KAIST, totally unaware of the technology and what it can do, and then I got introduced to this course [with] more than 150 people taking [it] The way this course was conducted was also very inspiring. There were a lot of interesting projects, hands-on stuff, and not just exams. So that was the course that inspired me to start my journey in that direction.


Was your thesis when you were in masters related to this field?

Yes. After I took this course, I thought that I should take more advanced courses. So I took Deep Learning for Robotics and Computer Vision, and then again another course in Computer Vision, which were both advanced courses meant for PhD/masters students. The core idea in these courses was that there were no exams, but mostly everything was hands-on. So for someone like me who wasn’t aware of tools like PyTorch, Tensorflow, and all those stuff that they use in machine learning, I was forced to jump right into the hands-on approach. Because of these courses, it eventually allowed me to do my thesis in the same topic. My master thesis was based on computer vision and deep learning.


When you went for masters at KAIST, were you very sure of what you wanted to do in your career, or was it shaped when you went to KAIST?

Before KAIST, I wasn’t sure what research meant and what academia meant. But while studying masters at KAIST, one interesting thing was in most parts of the world, at least from anecdotal experience, it seemed like masters was not a degree meant to do research in many parts of the world. It was only where you take a bunch of courses, and then a PhD makes you ready for research. It’s not expected to do research in masters in, for example, Europe. But here, it was a completely different environment that you had a P.I. and you were expected to do research. So I did get a taste of that, and I found out that this process of thinking about a problem, creating something, and pushing the boundaries of knowledge sparked joy for me. That is something I thought I would want to do for the rest of my life, if possible. This is especially true because before this time I had done a highly technical job for three years at the industry which, as I said, was related to control systems design, but I felt bored after a year because the learning curve was really steep at the start, but after a while it was just doing the same things again and again. But in research, you can tackle a new problem every day, and you can basically stay if you have the motivation. I found out that I had that motivation to solve new problems everyday.


So how did this help you decide on what you would do after you finish your masters? Were you very sure you would go for PhD, or back in the industry?

During my masters, I was able to publish my research in a top conference and that gave me confidence that I would be suitable to do more research, and the only job that fit into that description was PhD. That’s why at the end of my masters, I was sure I wanted to go forward with doing PhD. I’m very thankful to KAIST also because it gave me my first international experience where I met a lot of people in this conference, and I saw what it really meant to present your research at a forum. This conference had 10,000 [attendees] and was the top conference in computer vision and machine learning, and that gave me confidence that this is something I would want to do.


Which university did you go to for your PhD?

I went back to Pakistan after completing my masters. I had a year of mandatory service that I had to complete in my home country. After that was done, I applied to a bunch of places in the world, and my main focus was to apply in Europe. Therefore now, I am doing a PhD at the Norwegian University of Science and Technology.


You did mention though that you’re currently an AI consultant at Endress + Hauser in Germany.

Yes, so after completing my masters degree, I was also hired as a freelancing AI consultant. That is something that I pursue now. I have a company called Sarmad AI which is sole proprietorship, and I am able to offer my artificial intelligence consultancy services to companies who might be interested in that, but that is somewhat a side venture. My main area of focus was [previously] my mandatory service, and now [it is] my PhD, while I also do this consultation in my company.


In your PhD, are you focused on the same field?

Yes, my area of focus is deep learning and computer vision, but my department now is Computer Science.


What are some of your favorite memories at KAIST?

One of the things that I was involved in at KAIST was that I was an ambassador of KAIST for my home country Pakistan. That was part of the KAIST EE Liaison Program for getting new master and PhD students interested in applying to KAIST. As part of that program, I was sent back home to give a series of lectures and tell people more about KAIST. Those were most of the memories that I cherished a lot because I felt that I was able to contribute a lot towards my home country and my university, and KAIST was linked to this program. I remember in my second year, the program was already mature, and we invited about 80 students from different parts of the world, and I was asked to deliver a speech to these people. It was one of the first times that I was talking to people from different countries about KAIST, and that was something that I really enjoyed and cherished.


You also mentioned that besides being a student ambassador, you also participated in the KAIST EE Camp.

It was my professor, Hyunjoo Jenny Lee, along with other professors, who had a vision where KAIST needs more and more international students. Towards that goal, they noticed that people didn’t know a lot about KAIST. It was very funny when they shared to me that a student had an offer from Sejong University and KAIST with a full scholarship, but the student chose to go to Sejong while KAIST was higher up in the student rankings. So they were surprised about this, and whether anyone knew about KAIST. The vision of this camp was to send people home and tell the others what KAIST is. Because to be honest, when I applied to KAIST, I didn’t exactly know what KAIST is and what it represents, apart from a few friends. Therefore, it was really helpful that when I went back, I was able to talk to about 500 or 1,000 people in various seminars. Because of that, the interest increased so much that in the next year, we almost had double the applications in the EE department, and from there onwards I know a bunch of people I talked to who applied and got in. So I have a whole list of people who went to KAIST because they got to know the message through this program.


Did you face any challenges or problems while you were at KAIST, and how did you overcome those, if any?

I would say there are two aspects to the challenge: social and technical. Technically, after a gap of three years, I started my masters degree, and I felt that I was out of touch with the basics. I found it hard during the first semester to compete with the undergraduate and master students who were doing pretty good in these theory courses, and I had to work on my basics again to get into the whole thing. In the social perspective, I felt that a lot of the problems that I was expecting were not there, probably because there was no fixed timings that I had to follow on being in the lab. I was free to do what I wanted to do, so in that way I had a peace of mind. Secondly, I noticed that there was a huge Pakistani community on campus, and if I’m not mistaken even now they’re probably the biggest community there. So I almost felt that I was still in Pakistan. I kind of hoped for a more international experience, but when I went there, I felt like I was home again. So socially, I would say that I settled down really quick.


What about the technical aspects?

Technically, I would say that I felt that some of the skills related to theory and mathematics and the way the exams were conducted, I felt that it was something new to me, and I didn’t do very well at the start. Even the scores that I got in Deep Learning and AlphaGo, me and one of my other friends from Pakistan both got below average marks in the midterm. I wasn’t that worried then, because I realized that he was the top of his class, and he was really fresh, so he just came while I came in three years. The last time I opened a book was three years [before]. But then I realized that we could make up for all of that by performing well in the final project, which was something that I was able to work on and get a good grade. Then we felt that we have to change the way we study because we were not prepared for something like this. So from the next courses onwards, we were taking a different approach, and we’re really looking at the foundations and theory more for this kind of exam.


You told us that you learned about KAIST from one of your seniors and that time you had a great impression about KAIST as a university. How did your view of KAIST change throughout your stay? Did it become different when you were there compared to now that you are a graduate?

It definitely did because I didn’t realize how important a few things are in your life, and those are the people (professors) and the infrastructure and funding, and the name plus the research. Before this, I was only looking towards a few things like whether it’s Europe or Asia. That time I applied to a bunch of places, and I got a scholarship too in Italy. But I kind of chose myself to go to KAIST because of the great ranking. But when I told my friends that I am going to KAIST, they were skeptical on why I would go to Asia when you have the chance to go to Europe. So I had to fight off a lot of skeptics and believe that this is a great decision. I still feel that this is probably the best decision I ever made because of the opportunity that KAIST gave me and made me what I am today because of that decision.


In that regard, when you finished your masters degree, you went for your mandatory service back in Pakistan. Did you consider working in Korea or doing your PhD there?

The reason I didn’t continue was, one, I just went back to Pakistan for my mandatory service, so I had to make up my mind again. But also, I realized that there were a bunch of things that made me choose where I am today. The first thing is funding. During my masters at KAIST, we were allowed to increase the stipend and go all the way to 1.8 million KRW, which is about 1,800 USD, and for a PhD student it was about 2,500 USD. I just realized that for masters, someone giving you this kind of opportunity was an amazing thing, because no one gives you a masters degree and invests on you, since most people are not expected to do research. So universities don’t have that kind of incentive to fund you for a masters degree to a scale that KAIST is funding. Also, I didn’t mention it, but I had my twin sons during my stay in Korea, with my wife who previously had a masters degree. After we both wanted to pursue a PhD degree, preferably in the same place, we started applying and were looking for a place where both of us would get a degree, and because of the family aspect, it was also necessary to have that income level which would support me and my family. So from that perspective, I felt that Norway especially gives you salary in the sense that you are not paid the minimum wage. That’s why I chose to not go to Asia for my PhD.


If you could give advice to current KAIST international students, what would you give them?

Work hard and give in your best, but also take care of your mental health, because in the kind of business that academics are in and especially the times that we’re in, it can make you [feel] alone at times, and not everyone has someone to talk to. So it’s important that we as academics not only look after each other but also of ourselves, and to connect with people. And if you get some kind of rejection, it’s not the end of the world, so just keep on working towards the next thing.

출처 : The KAIST Herald(