The Job Search is a series for international graduate students.
By Anna De Cheke Qualls
Karthik Abinav Sankararaman (’19 PhD, Computer Science) grew up in Bengaluru, India – historically called the Garden City of India. More recently, it has also become the country's IT or tech capital. Despite the city’s rapid growth over the years, its green spaces and parks remain.
For most of Sankararaman’s childhood, Bengaluru was still a quiet, lazy town, but he could see the winds of change blowing.
“In 1999, personal computers weren’t commonplace in Indian households. Instead, there were these ‘cyber cafes’ in the city - essentially shops where people could go and use a computer on a per-hour basis. I vividly remember creating my first email account through Rediff, and being fascinated by the ability to send emails so rapidly. Over the years, I spent many hours in cyber cafes, playing computer games such as Dave, Wolf, Cricket, reading news and emails on Rediff, and also learning my first programming language, Logo. My partner in crime was my grandfather who was (and still is) fascinated by technology. He used to take my brother and I to the cafes where we spent what seemed like many hours in front of a computer screen. The remaining hours of the day was usually spent outdoors playing sports (well, mainly cricket) with friends in the neighborhood,” remembers Sankararaman.
Sankararaman pursued science like his father, an electrical engineer for Bharat Heavy Electrical Limited. His mother, despite finishing a master’s degree in Mathematics and Operations Research, chose to work at a bank so she could take her of her two sons.
“Growing up, we lived with extended family - where my father’s parents also lived with us. My mother’s parents used to also frequently visit. Given the diversity in people in our large household, I had a well-rounded and fun childhood,” recalls Sankararaman.
After high school, Sankararaman moved to Chennai, to pursue an undergraduate degree in physics at the Indian Institute of Technology. He quickly realized he had more interest in computer science.
“It was during my third year in undergrad when I was introduced to theoretical computer science. I loved it so much, that when it came time to do a PhD I decided to dig deeper into this area. My undergraduate days were a lot of fun - I was involved in making web and mobile apps for various events and organizations within the institute. Apart from my regular classes, this further developed my coding skills,” says Sankararaman.
In 2014 he arrived at the University of Maryland’s Department of Computer Science. While there, Sankararaman dabbled in many related disciplinary areas and opened himself up to meeting people from diverse backgrounds.
“This was the first time where I had the opportunity to work with people from various nationalities and walks of life. This drastically helped me improve my communication skills, and I also learned about cultural nuances,” says Sankararaman.
Just a few weeks ago, Sankararaman started a job at Facebook as a research scientist. He sat down with us to discuss his journey to industry, and lessons learned as an international graduate student at Maryland.
How did any childhood/young adults interests translate into studying computer science?
Looking back, I could point to various events in my childhood that I can attribute to my eventual focus on computer science. However, I believe I was doing most things without putting much thought into it and I was just “going with the flow”. It was only in undergrad that realized I hated physics, and had the opportunity to switch. It was then that I put serious thought into transitioning to computer science. I had some programming experience in high school and more or less enjoyed it – especially the part where we had to solve tough mathematical problems using programming. Later, I realized this is what we call algorithmic thinking, and over the years have put more and more effort into algorithm design and analysis.
When you came to study in U.S. what were your perceptions?
I had seen research originating from institutions in the U.S., but didn’t really know the process by which the results were produced. But it was evident that some of the most influential results in computer science came from American researchers. So, maybe U.S. universities had a “secret recipe” for producing massive amounts of research. Within a few months of arriving here, however, I realized that this university operates fundamentally in the same way as those in India. I did have to get used to a few cultural differences, such as calling everyone including professors by their first name, as well as taking initiative for things you care about if you want them to happen. But, the classes and research were pretty similar. It was just that most of the people who were good at this stuff tended to eventually immigrate to the U.S. I was clearly suffering from a “sampling bias”.
How is computer science perceived in your home country?
As I mentioned previously, my hometown back in India has become THE hub for computer science within India (and arguably within South Asia). In fact, outside of the U.S. it is one of the few centers where there is a high concentration of tech professionals. Thus, computer science is very highly regarded and sought after. For me, the main reason for coming to the U.S. was the availability of research opportunities, as perhaps compared to other parts of the world. Given that India is a rapidly growing economy and that tech companies are heavily investing in India, I anticipate that our research opportunities in India will grow dramatically in the future.
On a personal note, it was challenging to come to a new country/continent from a completely different culture and being far from home. But I immediately adapted to the circumstances and eventually started enjoying it.
How did you prepare for your non-academic job search?
In computer science a large chunk of people end up going to industry after the PhD. Thus, I was extremely fortunate to be in a similar circumstance as I started to explore career paths – with little preparation I could pivot to industry very easily. Also, with the plethora of former PhDs in the tech industry, it was not hard for me to connect with people and find out about career opportunities. The other thing that helped me was internships. I interned at various tech companies during my years at UMD. I interned at Adobe, IBM and Microsoft - each giving a different flavor of research in industry. These internships helped me drastically in both understanding how industrial research works in but also helped me to network.
What experiences as a graduate student (beyond your research) contributed to finding a position at Facebook?
As indicated, the primary influence during my graduate school years was internships in tech companies. It was also through my intern mentors that I connected with other people who had potential opportunities for me. The alumni network at UMD also helped a bit in connecting with people. Given that UMD has a large CS graduate program, it was not hard to find someone who worked in a company I was interested in. Computer science is in this golden age where there are lots of non-academic opportunities; in fact, through the abundance of opportunities many future employees suffer from the paradox of choice. Conferences were also a good place where I met people in industry who had either published papers or organized workshops or both.
What does networking mean in your home country? Any comparisons to the U.S.? How did you navigate networking as an international person?
The basic concept of networking is the same in India as here even though the ground level practices differ. In India, the main form of networking happens when someone speaks on your behalf or makes an introduction. In particular, it can happen that your advisor or a professor you worked with speaks on your behalf with their connections. On the other hand, here in the U.S., you initiate the conversation by reaching out to people. This was bit of a cultural shift for me initially, but over the years I embraced this and started reaching out to people more proactively. Once I realized this cultural nuance, it helped me network much more comfortably.
What advice or tool worked well as you searched? What did not work?
The single biggest tool that worked for me was the “cold email.” I was surprised at the number of people who were actually willing to connect me to others once I reached out to them and stated my needs. In general, I realized that if I was interested in a position, the best bet was to email someone who is related to that position and ask them if I am a good fit. Usually, many of these conversations converted into an informal interview where the person invited me for a talk or a formal onsite interview. The one thing that did not work was to meet people in conferences and ask them if they had open positions. At conferences, people are usually there to learn about research, and as a result, none of my overtures along these lines ever converted to a job lead.
Does a Facebook research scientist have a specialized skill set?
One of the skill sets necessary is a research background in machine learning. The other skill set they look for is strong programming skills. Usually, many research topics in machine learning require good programming skills as a pre-requisite. Since I worked on the mathematical aspects of machine learning, I did not code much in my PhD. However, I had enough coding practice in my undergraduate days that I could adapt if I had to. As well, I spent a few weeks refreshing on my coding skills before interviews.
How did you work on your CV/resume? What was the process? Who looked at it?
My advisor asked me to maintain an up-to-date CV and website starting with my first semester here at UMD. I took this advice seriously, and it helped me a great deal during the search process. I already had on-hand a CV that I had been refining over the years (with timely updates) through feedback from my advisor and other people. I also studied the CVs of recent assistant professors and people who joined the industry.
Career aspirations? What do you hope to do long term?
My long-term goal is to have an impact on multiple fronts: technical via solving important and hard problems, social via papers, talks, teaching and mentoring, and societal by working on problems that ensure that the algorithms I design work for everyone in every situation.
How is a U.S. work experience or education perceived in your home country?
U.S. work and study experience is still highly valued in India both in the tech industry as well as Indian academia. In fact, much of the rapid progress in India was a consequence of many Indians with a U.S. education going back to India and adapting their craft to the Indian context.
Where did you actually obtain work? What was the process like? Any surprises?
I will be joining Facebook as a research scientist in Menlo Park, California. I understood the process pretty well before I joined, so there weren’t many surprises in the process. I work on many problems that are derived both from products/infrastructure within Facebook as well as from the broader research community.
How do you continue networking locally? And globally?
Locally, I network by introducing myself to people and grabbing lunch/coffee with them to understand what they do. Globally, I network by attending conferences where many people from the academic/tech community show up.
What do you think an India culture and/or mindset (or approach, if such differences exists) can add to your work?
One of the ways in which growing up in Indian culture can help my work here/in the future is that I bring a perspective about the kind of problems I believe are important. Increasingly the problems computer scientists are solving is heavily impacting all aspects of society. Thus, having people come from different parts of the world adds flavor to both the type and the potential impact of collective problems.
A lot of young people are deeply interested in global social issues (equity, social justice and diversity for example) and I wonder if you have those interests as they pertain to your work?
A part of my research work deals with algorithmic fairness in various decision-making systems, such as ride-share and ads. Algorithmic fairness is an emerging and broad field of study where the goal is to understand how decision-making systems, especially those that are personalized, also lead to unfair outcomes. Personalization, such as recommending what movies to watch or what items you may like, in its current form is brittle and amplifies many of the implicit societal biases. Thus, the goal of parts of my research is to quantify, measure and eventually mitigate these biases.
Any advice to other international graduate students?
My advice would be to keep an open mind and learn from everyone. I think this is an attitude that would serve anyone well irrespective of the field they are in. It is also important to interact and learn about people from diverse countries/cultures/backgrounds. It expands your thinking and helps you grow in this increasingly global economy.
More information about Karthik Sankararaman can be found here.
(Photo credit: Karthik Sankararaman)