Shubh Jaju, MT1
Interned at: University of New South Wales
In my second year, I wanted to explore the field of research but didn't have any specific research interests at that moment. To have some clarity and direction, I explored the various CS-related domains I came across and read about them if they interested me. There are a lot of research topics we don't know about in the second year, so I believe it is important to explore different topics and fields instead of just mailing professors in ML and Data Science.
I started mailing around mid-December during the mid-sem break. Initially, I sent a few emails but didn't get a lot of replies. As I realized that the process was quite luck-based, I started mailing every professor whose topics interested me. By the end of January, I had sent over 150 mails and was also in talks with a few profs. Interestingly enough, the professor under whom I finally interned had replied to me 20 days after I sent him the mail.
CV, Cover Letter and CGPA
CV is one of the most challenging parts for most students mailing in their second years as many don't have a lot of projects or other prior internships to write on their CVs. I was stuck in a similar situation, but it is important to realize that most professors do not expect sophomores to have done research earlier. For a second-year CV, more than your projects, it is essential to include your technical skills and the courses/labs you have done.
I didn't write separate personalized cover letters for every professor. I would change a line or two for someone whose research interests seemed very exciting to me. It's not practical to read the research papers of every prof you want to mail. Thus I also feel that personalized cover letters are overrated and largely unnecessary for second-year students.
I don't believe that CGPA matters a lot, but a good 8.5+ CG might give you an edge over others. Professors around the world get a lot of emails regarding projects from various students. What they look for is seriousness and interest and how well one has performed in his previous undertakings.
I did my internship at UNSW, Sydney. It was a research internship under a professor from the CS department. He asked me to read a couple of research papers on a topic to see if I would be interested in working and discussed it over a video call. He later offered me a remote internship on the same.
My project was to solve a combinatorial optimization problem called the Social Golfer Problem using SAT encodings. I had no prior experience in this field and read a lot of papers. During my internship, I first read the already existing work in the area and then built my idea of a new encoding and implemented it in a boolean language (which I learned from scratch). I started with a very basic model and then, after a couple of iterations, started getting great results.
It is imperative to maintain good communication with the professor in an online internship. Luckily, I had a great online experience as my professor used to have a couple of meetings every week, and I could easily get in touch with him through the mail, to which he replied instantly. He would also schedule extra meetings whenever I got stuck. Obviously, an offline experience would've been much more fun but it was great learning even in an online mode.
Takeaways and Future plans
Though my internship was a great learning experience, and I got a much broader perspective on many new things, I realized that research is not something I'd be willing to pursue in the long run. I am really glad to have had this experience. It helped me eliminate the option of going into research, which I was skeptical about earlier. Second-year is the best time to explore the various options and decide if one wants to do a research internship or not. Once decided, the next difficult step is to start mailing. Personally, I delayed the mailing process because I couldn't build a good CV or thought I had very little knowledge of the various fields to go for a research internship. It is essential to realize that these aren't the qualities one requires to be qualified for a research internship; rather, one should have the desire to learn and explore a topic of their interest.
Interviewed by: Jaskeen Singh