We got in touch with Viswesh N(EE’24), who is currently pursuing his MS in Electrical and Computer Engineering (Focus area: Robotics) from the University of California San Diego.
Q: When did you decide to pursue an MS? What motivated that decision?
Ans: Around 3rd semester end or 4th semester start, I figured that I really liked robots and robotics, and not just in a superficial sense because I had spent some time with them, and then I realised I would need to pursue higher education for two main reasons: A) a lot of industry jobs actually require a higher degree for you to be to able to apply for those positions, and B) I was not sure about a PhD, I still am not. So with these two reasons in mind, I thought I probably should get a master’s.
Q: You said you are still not sure about a PhD, so with the degree you are doing currently, will you find equal opportunities in both — the research side and the industry side of robotics?
Ans: Yes, that’s correct. So the program I am enrolled in, has a thesis capstone or a coursework capstone, and I have 6 months to decide what I want to do, and depending on what I actually figure out in the 6 months if I choose to finish a thesis or if I just want to finish the coursework and get into the industry.
Q: How did you decide on which colleges to apply to?
Ans: For me, it was really easy. A word of advice for everybody applying to grad schools — I don’t think the concept of a safety school exists. So if you are spending the time and money to apply to a school, just make sure it’s a school you will attend. That being said, I only applied to the top 10 schools for CS or robotics research. I was a bit more picky when it came to the labs. So if there was a particular lab in a school that caught my interest then I applied to that school.
Q: Could you elaborate on the lab-picking process? How does a lab matter when you are applying to a Master’s degree because the degree is from the University, right?
Ans: Right. The degree is obviously from the University, and if you are willing to go to the industry, it really does not matter. But if you are double-minded about a PhD, it’s really important that you pick universities where you have both options open and specifically make sure the program offers a thesis requirement or a coursework requirement. Since I was not personally sure about a PhD, I actually have joined a lab right now to conduct research and see how it goes. For me, I had to pick universities where I could find labs that were a good fit.
Q: The research that you will be doing in that lab will not be a part of your coursework, it will be an extra thing, right? Like it happens in KGP?
Ans: Not exactly. Just like how we have a semester system in KGP, which is two semesters in a year, most schools in California have a quarter system. So there are 3 quarters and a summer quarter in a year, which is basically for internships and other things. So there are effectively 3 academic terms of study. If you decide to do academic research, you can do it just for the sake of it, like KGP. After that, you can start clearing your coursework requirements using research credits. If your professor thinks you are doing good work, you can get him to sign and get it graded, sort of how BTP and MTP work. You can convert your research into credits towards your graduation.
Q: Hypothetically speaking, if someone enrolled in such a program wants to go into a PhD, it will be much easier to get admitted for a PhD in the same college, right?
Ans: It does happen but it is not something you take for granted, especially in the top CS schools. I know for a fact that there are some groups of schools that are known for taking PhD students only from other schools in that group. I think Stanford, CMU, UCSD and there are selected labs that actually do this. This is some research you will have to do before you apply to schools as this also varies on your research area. Honestly, it does help in getting into that school or any other school in that network for PhD.
Q: Let us shift the conversation further back in time. Most people randomly try out different fields in their UG, so how did you decide your area of research?
Ans: That is kind of what I did too. In my first year, I explored embedded systems. I was more of a hardware guy, so I spent about 6–7 months doing that. Then like everybody, I wanted to explore what ML was. So I did a little bit of basic Machine Learning using some of the courses out there. Then I switched to Reinforcement Learning. That is where I liked what I was doing the most. Currently, I am working on Reinforcement Learning and Foundation Models for Robotics. I would say AGV actually helped me a lot here. I joined AGV at the end of my first year. Even within robotics, I tried a lot of different things. By the start or mid of my third year, I narrowed down to what I actually wanted to do — RL (Reinforcement Learning) and Foundation Models.
Q: How did you decide upon going to UCSD? You would have applied to a lot of Universities, what was the one thing about UCSD that made you feel like this is the one I want to go to?
Ans: I applied to 9 places in total and got rejected from my top 3 choices, which were CMU, Stanford and UT Austin, and this (UCSD) was next on the list. I had offers from UCLA, UCSD, NYU and Johns Hopkins University. Choosing UCSD was a mix of 3 reasons: a) their coursework was the best, it was not as rigorous and had the flexibility I needed over the two years to research or focus on the industry; b) the second reason was the labs, there are 7–8 really good labs here that work on very similar stuff that I wanted to do; and c) the location — both in terms of proximity to the industry, and just being like a really nice place for living.
Q: How is life as a Grad Student? How is campus life there?
Ans: So the Grad students are not really having the “Tempo”. But I still keep seeing events now and then, it’s fun. The Hike club is really active here so it is something I have been doing.
Q: How was your application process? How did it start and how long did it take?
Ans: Fortunately, I took enough time because of two reasons: I really wanted to do Inter IIT Tech and Socult in my Final year, so I had that factored into my timeline that I would be spending a lot of time towards this. So I started my application process around August since I had this right after my Summer Internship and I requested my Professors for an LoR (Letter of Recommendation), sometime around September. That’s something I would strongly recommend that you talk to the professors who are willing to write you an LoR and they are willing to sign as many LoRs as you can. Suppose you are applying to 8–9 schools so ensure you get permission for as many LoRs as you need or you have 4–5 letter writers, so you can mix and match and get to as many letters as you can. So 3 is the standard for every school, some schools accept up to 4, but 3 is necessary and sufficient. With that being said, I’d say August mid is when I started writing the first draft of my SOP and shortlisting schools.
Q: What about exams like GRE, TOEFL, IELTS? When did you start preparing for those?
Ans: So in my case, I did not really give a lot of time for preparation. I’d say 2 weeks is more than enough for GRE or 3 maybe, if you are willing to stretch it. The Quant section is really just 10th-grade Mathematics, so you should be able to do really well on it. For the verbal section, it depends. You can probably look at past papers, check where you stand, and depending on that you can choose to give more or less time. TOEFL again is a really easy exam. It is just designed to see if you can speak in English or write in English. You should be okay with giving 2–3 weeks of preparation, 3 weeks, if you are more cautious about it.
Q: On a relative scale, how important are all of these: research experience, GPA, LoR, SOP and GRE/TOEFL?
Ans: I cannot emphasise this enough. Specifically for a Master’s, GPA is very important. Since you are not applying to a PhD program, a lot of committees are going to use GPA as a primary source of filter. It does become less relevant if you do apply for a PhD, or if you are applying for a thesis-based master’s. So the only schools I got into were where the decent programs were Thesis-based Master’s, which did not really give a lot of focus on CG. I’d say 9.5 or above would put you in the top bracket or Department rank 1, 2 or 3. Specifically for Stanford, they only take Department rankers. So it is very important that if you are targeting those types of schools, make sure you are in the top 5 department ranks, apart from that I’d say 9+ is a very safe number, or to the point where it really does not matter that much if other aspects of your application are good. 8.5 is the bare minimum and below that, I know a lot of folks who did not have any admits in the 8 to 8.5 range. On a relative scale of importance, if I had to rank them, GPA >= LoRs = Research Experience >= Test scores = SOP. About SOP, I have the opinion that it really does not matter much but it can definitely destroy your application if you cannot write it properly, but a stellar SOP by itself won’t get you in.
Q: You said course-based master’s and thesis-based master’s, can you explain the difference between the two?
Ans: Course-based master’s are specifically for the industry so a good example of course-based master’s would be Stanford’s MS programme or CMU’s MRSD programme. CMU actually has a lot of these programmes MIIS, MRSD depending on what you want. Thesis-based master’s are sort of more focused on research. It’s for people who are more or less sure that they want to get a PhD but don’t have the profile to get into the top CS PhD programmes so that would be the type of people getting into these programmes. Their cohort sizes are in general really small so if you check Cornell or UIUC or UT Austin, they have like thesis-based master’s where they actually have sort of small cohort sizes and they are very competitive and you’ll be spending about six months to one year just doing research.
Q: How did you manage the time out of your coursework and research work to prepare for the admission process? Or were you doing some active research in your final year?
Ans: Coursework in particular wasn’t very taxing to me. I was doing my BTP and I already had some progress on my BTP like over the summers so I largely presented that plus a little bit for my first BTP requirement which happens at the end of fall semester. As for other stuff, I did two Inter IITs so that’s where most of my time went. But then again, I think the commitment started somewhere around late November or maybe a little bit earlier than that, I did the mid-prep events so I basically had all that time leading up to November. As for coursework, most final year courses are really chill and I think it’s 2–3 courses per semester if you plan wisely and I also needed to take a humanities course to finish my requirement so I sort of kept coursework on the light side while focusing on this.
Q: Did you have any ongoing internships or something in the final year?
Ans: No, not at that time. So my summer internship, I was continuing that because it was a paper submission in progress like back from the summers but nothing apart from that.
Q: How was your summer internship?
Ans: Yeah so my summer internship was at the University of Toronto through Mitacs. I worked with Professor Igor Gilitschenski and it was nice. I think it really sort of gave me an exposure to research from a perspective that I hadn’t really gotten before. I had a lot of hands-on experience, something which was unfortunate for me since the first two years were online, so that was something nice about it and obviously I think the LOR really helped me in terms of other places where my profile was lacking.
Q: So Mitacs and DAAD, both have been stopped right, so do you have any recommendations for new UG students who want to experience research outside India but don’t know how to approach it?
Ans: So about this, I know a lot of my friends who used KGPF funds. So, you know, IR Cell rolls out these applications for … So the only thing I found sort of dicey about it was the timeline. So I would say if you’re very serious about going to US or Europe where the visa process actually takes some time, try reaching out to them in advance, don’t wait for FTP to sort of roll out those applications because they are willing to help you, they do have money to support you, it’s just that there is a timeline disconnect. I know a lot of people who went to Israel, who went to Australia and other countries where the Visa process isn’t as lengthy and they were able to use KGPF’s funds. So just mail as you would, do your cold mailing, try to write to professors and try to arrange for funds through KGPF.
Q: How would you define the experience of research in India and in foreign countries? How is it different?
Ans. So the two things that I noticed are the supervision style and the funding. If you look at labs abroad, there are two kinds of labs, the ones that are really really big with well-established grad students, so in labs like these you’re almost never going to meet your Professor but there’s like a very clean hierarchy setup, sort of like how corporate works, you know like, you’ll be reporting to a PhD student and he’ll have like regular meetings with the PI which you’ll be a part of as well. Or you could opt for like a more direct style where you have bi-weekly meetings with your professor. I think in India it is more like a sort of direct Professor interaction. Two is obviously the funding, there’s infinite money here and I’m not just making this up. There are companies like NVIDIA, DeepMind, which are pouring money into academia labs, so you’ll never ever run out of hardware or software resources. I think especially in a time like now where like having an H100 cluster is the norm, I think it’s very important to factor in the funding aspect of things.
Q: Does your UG degree matter to where you’re applying?
Ans. It didn’t really matter to me because a lot of schools in the US have ECS as a single department so I didn’t really face that bottleneck. But a lot of schools do mention that you have to be from an ECS background to apply. UT Austin was one such school. So just check before applying for any school. Another thing that you can do is going on LinkedIn and see if there have been people from other departments who have gotten into these programs. And this applies for any branch, it becomes a little more pronounced in core engineering departments where it’s very hard to switch to mechanical or something if you’re doing something completely unrelated. One way people mitigate this is through work experience, so if your major isn’t directly related to the course you’re applying to, a common approach is to work for 2 or 3 years, bridge that gap, and then apply to master’s programmes.
Q. What are some factors that make one’s application stand out from the rest of the crowd?
Ans. Obviously, for master’s, a stellar GPA i.e. top 3 of your department plays a huge factor, it puts you in a different category. Published papers that you have penned really matter, for thesis-based master’s. Really good letters of recommendation from professors who are well known or who have collaborated with the university you are going to. I would say these 3 things could really make or break your application, even if other areas are lacking.
Q. How are you managing expenses for your master’s programme?
Ans. I have taken a loan initially, at least here, the system is if you get a research assistantship or a teaching assistantship, your entire tuition gets waived and you get paid a stipend. There are very few schools that do this. That was also a factor when I was picking schools, so I didn’t want to apply for very expensive schools like the Ivies basically, so I mainly just applied to public schools with the exception of CMU and Stanford, You can look for master’s programs where your tuition is going to be partially funded, or at least have the option to procure funding, so that you don’t have to take a loan for the entire duration. For what it’s worth, if you are going to be entering the industry, you will be fine taking a loan, like it is not the end of the world.
Q. As an alumnus, could you give some words of advice to current undergrads of IIT KGP?
Ans. Try to explore as many things as you can, not only in an academic research sense, try to play a sport or two, try to join socult societies, and really put yourself out there. There’s one other thing I would like to tell you, though it may sound extremely cliched, the friend group you will have in Kharagpur is probably going to be unmatched going forward in life, most of your relationships are going to be professional. So keep your friends close as it matters more once you graduate college. Apart from that, plan your time wisely, don’t slack around. I wouldn’t give you advice about attending classes but be a little sensible when it comes to midsem and endsem prep. Even one week of solid effort will put you in the top one percentile if you are sensible with your time. And definitely like do play sports if that’s your cup of tea, do get involved in socult, because these are things you won’t really get to do once you graduate.
