Vinayak Bansal (CH1)
Interned at Max Planck Institute for Dynamics of Complex Technical Systems
Domain: Model Order Reduction
The primary driving force behind this decision was that I began with machine learning in my second year and developed an interest in it. There was no specific reason for choosing machine learning per se; I just wanted to enter computer science, so I chose it. I am considering switching my program, and I would like to pursue a master's in CS after my Chemical Engineering. As of that time, though, I didn't have proper experience in CS, so I thought a research internship would provide valuable insights and help me clear my head on whether I should go for a master's and beyond. A research internship would also look better on my CV than a corporate one for pursuing higher studies.
Preparation and Process of Application:
I prepared for the research internship by delving into machine learning through online courses, with a particular emphasis on fundamental topics. I gained practical experience by working on coding projects that handled datasets and applied machine learning concepts. To secure a research opportunity, I proactively contacted professors within and outside my field of study and was fortunate to land a project related to machine learning applications in healthcare. Additionally, I obtained letters of recommendation from professors who could attest to my commitment and capabilities, showcasing these experiences on my CV to demonstrate my genuine interest and proficiency in the field.
I was very late in the mailing process. I started cold-emailing professors from around mid-December. I created a base mail template pertaining to my skills and areas of interest, personalised each mail according to the professor's past works and some research papers they wrote, which piqued my interest. I scheduled these mails to reach their mailboxes in the morning and refrained from sending them on Monday or Friday. The logic behind this decision was that on Mondays, all the mails from the weekend pile up, and on Fridays, professors are usually in haste to wrap up all pending work.
(How many professors did you mail?) So, I was lucky in this scenario. I had to just compose 100 personalised emails. I have heard about people who have sent over 800 personalised emails, which means that they visited each professor's webpage, extracted information regarding their work, and then wrote customised messages to each. I got 7 responses: 6 responses were negative, and only one was positive. In my opinion, the outcomes of this entire process are very probability-based.
After receiving a positive response, I visited the professor's website again and thoroughly analysed two recent research papers that she had authored. I also prepared some slides on the projects that I had done. During the interview, I first presented my work and we talked about it. The professor gave me a deeper glimpse of her ongoing projects and then I asked her the questions I had when I read a paper. She happily answered them and then gave me some work. It consisted of reading and understanding a paper with code, and then applying that to another problem. This process lasted for another 10-15 days and then based on my performance, she selected me.
The Internship: Apprehensions, Experiences and Takeaways
In preparation for my internship, I learned some basic German phrases because English was not widely spoken in Germany, and having a translator was essential for communication. I also learned basic cooking skills, as eating out daily was not financially feasible. In terms of my internship projects, I worked on three projects during my first internship. One of the projects focused on using machine learning to approximate differential equations, aiming to make complex algorithms with high time complexity more efficient. It is crucial for applications such as mobile apps, where solving differential equations in real time is impractical. I also worked on projects involving data classification from rat brains and using machine learning to approximate transfer functions in process and control engineering. My internship experience was challenging because it was my first international trip, and travelling alone. Still, I found support from the network of fellow interns living in the same guest house, which helped alleviate most emotional and mental challenges. I was living in the institute's guest house, so I didn't have to look for accommodation, and there were five other people with me to work and hang out with.
Express your interest in the field of study and express a desire to learn and gain insights. Professors value enthusiasm and interest in their work, so ensure your application and cover letter reflect your passion for the field. This can lead to opportunities to work on research projects.
Consider the timing of your application. Avoid sending emails during the Christmas holidays when professors are not actively checking emails. Try to schedule your emails to arrive in the professor's inbox in the morning. Also, try not to send emails on Mondays or Fridays, as these days may have more crowded inboxes due to weekend and Monday morning emails.
While a higher CGPA is a strong advantage, it's not the only factor. Some individuals with lower CGPAs have secured internships, but having a high CGPA increases your chances. So don’t be demotivated, You never know. Besides CGPA, projects and relevant academic background can also play a significant role in securing an internship. Highlight your projects and academic achievements. If you have other factors that can make up for your CGPA, don’t hold back.
Obtaining solid Letters of Recommendation (LORs) is crucial for your application. Building relationships with professors and seeking LORs from those who can attest to your academic abilities, work ethic, and dedication to the field will enhance your chances of securing a foreign research internship. These letters provide concrete evidence of your qualifications and character, making a compelling case for your suitability for the position.