Preparation Process:
I started the preparation in late July. I found the tests to be of moderate difficulty level. There are usually a few questions related to quant, dynamic programming and data structures. The questions related to ML and regression are basic, and any course related to regression is usually sufficient to crack them. Some questions are like Brain Stellar's, and others are mental ability type questions. I mainly revised my knowledge of ML and data structures for the interviews.
Application Process:
I applied for some consulting and data analyst profiles as well. After the data analyst interviews, I realised that the profile did not align well with my future goals, so I did not pursue it further. Later I applied for Mastercard and got the offer, so I decided to take it up. I applied for limited companies only after day two.
Interview Experience:
I had two rounds of interviews. The company evaluates your logical thinking and problem-solving skills in the first interview. In my interview, the interviewer mainly discussed my project in detail. He also asked me some open-ended questions related to the project that required a good understanding of the basics. The second interview was a short one. The interviewer asked me some basic questions related to ML and a few about finance since I mentioned that I was preparing for CFA.
What made you stand out?
I had done a project related to machine learning, which primarily made me stand out. We had an extensive 20-30 minute discussion on my project in an interview. I was able to explain the intricacies of my project with clarity and was able to give well-reasoned answers to all the interviewer’s questions about it. The interviewer was very impressed by this.
Work at Mastercard:
I was fortunate because I got to experience both business and research aspects of Machine Learning. I did application-based research on the ML model of a recommendations engine for a food chain company. We then used the insights from that project in the primary research. I also wrote a research paper based on the model I made during the internship.
I had a great experience working at Mastercard. Initially, I was not very sure of Machine Learning as a career prospect and was also exploring consulting. My core interest was problem-solving, not primarily Machine Learning. However, I enjoyed my work at Mastercard so much that I accepted the PPO and will be continuing to work there.
At Mastercard, I learned that machine learning is not just about coding. Coding is just one aspect of the work, and it also involves a lot of brainstorming and innovative thinking.
Work Culture:
I felt that Mastercard had a perfect work-life balance. The work used to happen only on the weekdays, and I did not have any workload on the weekends. The team also made efforts to include the interns in the projects and mentored me whenever I faced any problems. In the research paper I published, I was credited as the first author which I have heard is rare in most companies since usually the manager is credited as the first author.
Essential Skills:
I did not acquire my knowledge of machine learning directly from college courses. Nonetheless, I have chosen it as my career path. So the courses at IITD did not play an essential role for me. The most important skill I learned in college is communication. Even as a beginner in ML, I was able to interact with the team confidently. I always had the confidence that sooner or later, I would be able to finish my tasks if I put my mind to them.
Online vs Offline:
Overall, my work did not require much on-site work. So I did not face any major problems in remote work. I initially faced some configuration related issues that were avoidable had the work been offline, but they were eventually resolved. I was also able to network with the team well on video conferences; however, the absence of in-person interaction was certainly felt during the general meetings among the team.
Memorable incidents or highlights:
Going into the internship, I thought that I was good at making presentations. Despite that, I realised that the longest meetings I had were not related to my work on the ML models but rather the ppts that we had to make. This came as a surprise, and I learnt that presentations in the corporate world are very different from what we usually see in college.
Interviewed by: Nikhil Gupta