“And I knew exactly what to do. But in a much more real sense, I had no idea what to do.”
― Michael Scott
Hello, I am Omkar Ghugarkar, a third-year student at IIT Bombay with majors in Chemical Engineering and minors in Data Science. I am from Pune. I am into ML with an interest in Computer vision and like to create real World applications. I am also a huge fan of Cricket, and pre-covid used to enjoy playing cricket, but now my hobbies include binging Web series, playing online games, and watching football matches. I also invest in the stock market, although I am just a beginner.
What to do in second-year breaks?
To make oneself more credible for future internships, jobs, etc., one would indeed require skills and knowledge other than the courses undertaken. The breaks are the best time to acquire those skill sets needed.
What important choices does one need to make in second-year breaks? It starts from the very root. Which field should I focus on? Core, Non-core. If it’s non-core, then what…. And the list goes on. Second-year winter break probably is the best time to narrow down your interests and pursue one or two things in which you surely can see yourself progressing. You surely don’t want to be “Jack of all, master of none.” What you should try is to master atleast one.
How to make the most of the second-year breaks?
Once you have narrowed down the interest, what are options one can perceive? Learning and mastering skills, project with professors within the institute, research intern, or an intern at a company. All these aspects require different levels of skill sets and also depend very much on the individual. The main focus should always be on learning.
“Your only competition should be with the person you were yesterday. Improve everyday, so your past self doesn't win.”
My second-year Winter intern
I also work as a Senior Computer vision Engineer at SeDriCa, UMIC. Having explored and pursued Machine learning and Deep Learning since the first year, I was in good shape to explore the real-world problem. My main motto was to get hands-on experience and do something more significant than a toy problem.
The problem now is who will offer me an internship. There are very few options from the PT Cell for second-year students. Pursuing ML and being in Chemical Engineering does not help. I once heard a talk and learned about the website, which offers interns for some newly established startups. https://angel.co/. Internshala never looked ok to me because there is one post, and you can see there are 200-300 applications.
I applied for around 10 ML jobs and got a response from only one. Being just a sophomore and non-CS background simply doesn’t help. I was fortunate enough to get at least an answer.
The application procedure was simply to apply via resume and fill in the extra details if asked. The interview was based entirely on my resume. I was asked to explain all the structures used, losses, and optimizations for the various projects on my resume. And guess what... Everything went well, and I was offered the position.
The name of the startup was “Machine Learning Studies.” The idea was to make recorded lectures and videos, which are 2D in nature and turn them into 3D. My task was to work on 2D faces. Everything started from absolute zero- from gathering datasets to cleaning them, preprocessing them, and finally, learning a model for the required output. We did implement some papers from scratch, and there was intensive learning. I learned about the pipelines that are followed in real-world applications and implemented some research papers from scratch. Last but not least, the satisfaction you get when your thing works.
My second-year Summer Intern
It got this intern through PT cell. The company’s name is “Vedantu'', and I worked as a Data Science intern (yeah, I was not teaching XD). The structure was very standard—a coding and aptitude test followed by an interview. The coding test consisted of two questions. I was able to do one question completely and passed some tests of the other. The aptitude test was somewhat easy. Fifteen students were shortlisted for interviews. And again, like the previous interview, it was based entirely on my resume. I was also asked about my work experience of the earlier intern and some ML questions that were very close to what I have done. Out of the fifteen students, three were selected.
It was a three months internship. The earlier half was based on the Computer vision aspect and after that on audio. We developed a modulation system that can output the quality of speech without considering the discussed subject matter—this is a combination of many Deep Learning models and audio techniques that I eventually learned over the intern duration.
This internship was vital because it was a matter of discussion in the interview for my third-year intern and played a significant role in securing an internship at Daikin, Japan.
What if I had not interned (My Backup Plans)
When it comes to learning new stuff in ML, I always had the SeDriCa work passing by. Along with this, I had a plan to take on projects with a Professor within the institute. iSURP and SURP are excellent for this. One best thing about the ML field is that, once you are done with the basics, you can always read research papers and implement them from scratch. And if you feel you are missing something, go back to basics!
To summarise, one should utilize the second year breaks to gain a good amount of knowledge in a particular field. It might be through interns, through self-learning courses, projects with professors.
I hope you have an excellent Sophomore year at the institute:)