Hetvi Jethwani - Mitacs



Topic and Field:

I signed up for a project called 'Summarizing Machine Learning Model Predictions,' and it sounds like it's machine learning, but that's not it. It was a problem in Theoretical Computer Science, specifically Algorithm Design for Database Summarization. It involved more math than I had expected, which was really fun.


Why Research

I was quite unsure about research when I joined college; all I knew was that I liked math, and it was what I wanted to do. I think I had the research kick from before college, especially when, as a kid, you find all things that scientists do mysterious and fascinating, especially in movies (Dhoom 2 was my STEM awakening). But after I joined IIT, I was pretty clueless; I was doing various things and projects, which turned out to be immensely helpful and fun. I also realized that I do not like competitive coding at all and was repulsed by the coding tests one has to give during the training and placement processes for industry jobs. It seemed mechanical to me; some people, of course, might enjoy it and approach the problem like it's a puzzle, but that wasn't the case for me. Something that really pushed me towards research happened in my second year. As a part of the Physics and Astronomy Club, I worked with 2 or 3 alumni on a literature review of 10-15 papers in Computational Neuroscience. I spent a lot of time reading and learning to read research papers, and I loved that. The range of the papers was huge, and even though I didn't understand a lot of it, I liked and enjoyed it. Wanting to try research, I took up a project under a professor and started applying subsequently.


Finding out about the internship

I knew about Mitacs through one of my seniors who had gone for this internship in her second year (Earlier, it was open to second-year students). I was really worried about not getting a corporate internship because at IIT in general, even if there's nothing to worry about, you still get conscious of your decisions when they deviate from the norm; everyone around you would have an internship or would be enjoying at least a little bit of success. It's like that with research, but my professor reassured me and told me not to worry as there are good research opportunities in the industry as well. The really good thing about Mitacs is that they pay you, which is not assured, especially in cases of cold mailing.


Application and Selection Process

I think eligibility, more or less, is students having 'one to three semesters remaining in their program as of Fall 2022'. The exact eligibility is on their website, and they're very responsive over mail as well regarding any queries.


You first fill out a standard application form; there are sections about your personal achievements etc., but the other important part of the application is choosing your projects. The projects are from all departments, from mathematics to humanities. Some projects explicitly state that they need a person with a specific skill set, e.g., someone who knows Java. You can fill out a preference for a maximum of 7 projects, and after that, it's like JoSAA but with interviews. Most professors take interviews; some don't. It's really dependent on both the professor as well as the volume of applications of the project. In my case, I had interviews for projects with priority ranks 1 and 2. I got an acceptance in the first one, so I was then out of the process for other projects. And say if I had gotten acceptance for the third one, it would, for sure, mean that professors 1 and 2 had not accepted me.


Regarding the interview part, I believe that I was able to clear it as I had explored a variety of stuff earlier (the 1 2 3 4 project, Computational Neuroscience for PAC, a side project in making an app in the middle of a hackathon, and a subsequent project under a professor) and it helped that I had applied for a theoretical project that required more mathematical background. The first question I've always been asked is what is this mathematics and computing degree and how is it different from computer science; in a theoretical sphere, it helps me stand out as, by virtue of coursework, it has made me more rigorous when it comes to approaching problems. The professor asked me 3-4 questions, and one of them was to talk about a problem that I really enjoyed solving even if I couldn't reach the solution. I think he could see my enthusiasm for the problem; it didn't matter much if my skill set was specific or not, I think I showed my enthusiasm and the capability to be driven to the professor by that question as well as accept that I wasn't able to solve a problem and did some jugaad to make it work. It showed that I was enthusiastic, could describe the problem to another person, and accepted that I couldn't properly solve it, which is quite difficult for people but simultaneously important in research.


Expectations Beforehand

While applying, we had to sign this contract which basically said that we'd come to Canada, but then within two weeks, they sent us an email that it wouldn't be possible because of COVID-19, so that was unfortunate.


What I had heard about Mitacs before was that it's not possible to make a significant contribution in such a short time, and it's mostly just coding work. But fortunately, I enjoyed the work and wanted to take it to a position where some other student could pick it up. I found the UI building part slightly challenging, and had not expected UI development in Java to be that difficult.


Overall Experience

It was really good, even if it was online. They used to have weekly meetings and didn't micromanage me at all. They weren't demanding daily updates; they gave me the space to read, and I was a bit slow at work because the problems we were thinking about weren't apparently trivial either. They were responsive over email as well, and even though it was remote, I felt that I had the time and space to actually think about things, read up and ask questions since the meetings weren't that frequent. I really liked the project, and it was kind of unique in the sense of having two supervisors. One was a professor at the Canadian university and the other a researcher at a lab in the USA. Now that I'm applying for M.S. and Ph.D., I have inputs from a researcher in academia and a researcher in the industry who are very helpful.


Takeaways

I think my biggest takeaway would be that it's good to be vocal and ask questions, even if you think it's trivial. It's important to ask questions to the people you're working with so that you don't get stuck at certain points.


The Relationship for Future

I think the biggest help for me with respect to this internship is the connections I've made with not one but two very helpful supervisors. I think the academic relationship I've built with them is quite valuable; both of them are in a database field, and even if that's not an area of my interest, just knowing and connecting with such brilliant people while knowing how they solve and approach problems is precious to me. Knowing that I can email the two of them if I ever need help in the academic sense of things, e.g., if I wanted to connect with another professor in the same university, I can email this person and ask them to connect us and so on.

 

Interviewed by: Priyanshi Gupta

164 views0 comments

Recent Posts

See All