Apar Ahuja, CS1
Interned at: New York University
Application Process
I started mailing professors around August. I never had any specific field in mind and chose the professors according to the projects they had worked on and were working on at the time. I chose the professors using a website called csrankings.org. It has a list of professors and their universities sorted according to their countries, research fields, and the number of citations. I found this process to be much more straightforward than the tedious process of going through university websites and finding professors and their contacts there. Then came the process of mailing all the professors. I had one primary mail, which I altered according to the professor I was mailing and their research fields. It ended up saving me a lot of time and effort too. I was finally offered an internship at New York University (NYU) in the field of computational biology.
The Internship
My internship was online, so I worked from home. The professor was an IIT-KGP alumnus himself, which made us easier to relate to. He was able to understand what I knew as well as my inclination. In my opinion, it would be a good idea to target IIT alumni while mailing professors as well. However, I didn’t interact much with the professor. Inav, a master’s student, worked with the professor on this project, and most of my interactions were with him instead. This proved to be an advantage since the sessions weren’t too formal, and since Inav had come back to India due to the COVID situation, the timing wasn’t an issue either. They took my feedback into account, considered my ideas, explained to me the problems if the ideas didn’t work, and when I thought of a good idea, they let me implement it too. It felt nice to be able to implement something that I had thought of myself. Since we were a group of 3 interns working alongside Inav, the momentum never suffered, even if one of us had a slight delay with the work.
The project we were working on is called Episimmer. Essentially, we were developing a simulation environment for the spread of diseases. It aims to help colleges and schools open up during the pandemic time and recommend strategies and policies about lockdowns, vaccinations, pool testings, etc. We tried to make a system that college authorities could use.
For the first half of the internship, we developed the simulation environment. We worked on adding various features to it, like mathematical recommendation and visualization features. The second half was applying reinforcement learning to recommendation features. Everything else was essentially the implementation of ideas. I actually learned reinforcement learning for the first time through this internship. I was initially apprehensive about learning it, but the Inav just told me to study for two days and was asked to implement it on the weekend. It felt like MTL courses were actually being used in real life :P
Episimmer is currently a public platform and is used by IIT Jodhpur for recommendation and policy. It’s also been used in IISC Bangalore for its epidemiology course. Version 1.0 of this platform was publicly released a week or two before my internship ended. So, I was a part of the finalization process and helped fix bugs, clean up code, etc., which felt great.
Takeaways
This internship helped me learn a lot of new things and was, all in all, a very fulfilling experience. I recommend that everyone should try for a research internship. There is no such GPA criteria or anything like that, and with enough effort, anyone can score a good internship. Good luck to those applying!
Interviewed by: Sana Kumar
Kommentare