Himanshu Wadhwa - CISPA Helmholtz Centre for Information Security, Germany
Updated: Dec 30, 2022
Brief intro about the intern
My internship was an offline internship in Germany for” Cisca Helmholtz Centre for Information Security”. It was in the domain of typical machine learning. It was more theoretical, and they gave me an independent project. The internship was for two months, but I have continued working on the project to date to get a paper published.
There were two significant factors that motivated me to pursue a research internship. The first factor was peer influence, and the second was the kind of lifestyle I wanted for myself. I didn't want a typical coding or 9-5 job; I wanted to join an institute to do research. In the long term, I want to get into industrial research. So I gathered the information I needed to pursue industrial research. Typically a master's is required for such a job as you don't get such in-depth knowledge of the topic in a bachelor's degree. Hence I decided I wanted to do an MS, so I reached out to seniors who had similar interests, and they suggested a research internship, where I could work under a professor and get a paper published. Doing a research internship looks good on your master’s CV, and it's a great opportunity to get an LOR. These are not only beneficial for MS but even if you want a job later on; it looks good that you did something in your summer instead of sitting idle.
I started making my CV in October, as suggested by my mentor, but I couldn't complete it as I was involved in other activities. Then majors came, and I got occupied with that. Just after my majors, I started mailing. I made different versions of my CV and got it reviewed by my seniors till I reached the final iteration. November and December is the ideal time for mailing after January; you start getting negative replies, as professors have already accepted someone. Hence it would help if you got your CV made and reviewed by seniors till then.
You have to be selective in mailing. The QS ranking site will give you information about the professor and their research to see which suits you the best. After finding professors, go to their web pages to see what research they are currently involved in, and whether they want an intern. Try to reference something from their research while mailing. Also try to look for personal websites since that can give you an idea of the kind of mail the professor wants. Sometimes professors state on their page the subject that should be written in the email while applying for the internship; otherwise, the mail gets lost or ignored. Read about the professor's interests and align your cover letter. Personal web pages are better than faculty web pages as they are frequently updated compared to faculty web pages.
There are no CG criteria to apply for summer internships if you apply directly. But there is a CG criterion if you go through programmes like the EPFL summer intern programme, Mitacs programme, etc. When it comes to getting accepted the more the merrier usually works.
Preparation for internship
During my first year, I tried participating in different clubs. I was a part of AXL8R, robotics and the debating society. From these clubs, I got a start and learnt that you need to do projects for your CV. I also participated in many hackathons and took up a summer project in Rutag. In my opinion, the most critical task is finding a good group that helps you grow and gain experience.
During my second year, I mailed IIT profs to get projects and got one under professor Manan Suri. I also did self-projects online, apart from the two I did for COL106, a coding project and another in research signal processing.
It would help if you started exploring from the first year. It would help if you think about your interests, not just money, and whether you would like to pursue them in the long run. You should always analyze the advice given to you and follow the one that best suits you, as everyone is looking for something different in life, so there is no one way to do things.
Experience -Neural network pruning/optimization
I interned at "Cisca Helmholtz Centre for Information Security" in Germany. It was an offline internship and was about Neural network pruning/optimization, a machine learning domain. It was a highly theoretical one. The internship was for two months, but I have continued working on the project to date to get a paper published.
My role in my internship was to be an independent researcher, and I used to interact with the PG students and professors occasionally. I was part of a group that sat in the office that worked under the professor. They gave me an independent project and the research paper it was based on to get an idea of the project. I read the paper, and it solved my doubts and difficulties. After reading the paper, they asked me whether I was still interested in it, and I said yes. It was simple and was based on experiments under changing values.
The work hours were relaxed, and it was never a compulsion to go every day. Due to covid, there was a limit on how many people could sit in a room, so it wasn't compulsory to go there every day. Since my housing was very near, my professor also came twice a week to meet me.
Living expenses were not sponsored, but they gave a decent stipend to cover all costs; the rest depends on how you want to live. I saved some as I only spent a little on traveling to places.
I made some great friends there. There were some Indian students on the campus. The research institute is located near an already established campus with famous CS research institutes like Max Planck. There was a well-established center for international students which hosted parties weekly or bi-monthly; I also went on tours, to cafes, bars etc., with them.
Advice for juniors
Don't pursue a research internship only because everyone is doing it, do it only if you want to go into research. Make your own decisions based on your interest, and don't get influenced by others. If you are going there only to see the culture and interact with people, then a similar experience can also be availed through Forex, which might be even better. Since the professor invests in you, they will also want you to do some work. If you do not know stuff and go there only because of peer pressure, it might be tough to manage as you are expected to know things you put on your CV beforehand. Hence do some courses from Udemy or Coursera to build your CV. Also, talk to the professor and ask them what expectations they have from you, and then you will realize whether you are fit for it. Research interns will only be of value if you do some work and not just if you go there to enjoy. Because when you put them in your CV for whatever domain, you might make a shortlist based on it, but the interviewer will always ask you how you have contributed and judge you based on that.
Interviewed by - Rudrakshi Singhal