23/09/2026
“Research was never part of some grand plan I had made for myself as a school kid. I chose science in high school because I loved mathematics, physics, chemistry and biology, and eventually found my way to computer science. During my undergraduation studies, I fell in love with the subject. But it was only when I joined IISc for my Master’s that I began to understand what research actually meant and discovered the beauty of deeply exploring challenging problems and uncovering new ideas.
Professors there taught us to think, to sit with a question, challenge assumptions and find ways of answering something that did not already have an answer. After my Master’s, however, I didn’t immediately pursue a PhD. I joined Xerox Research in Bangalore. I had a permanent, well-paying job, was doing research and writing papers, and life was comfortable. Then came the slightly irrational question: if I am already doing research and writing papers, why not do a PhD on something concrete? There was one very rational reason not to; moving from industry to a PhD would financially be a bad decision.
My parents understandably asked why I would leave a secure job. I asked myself the same thing many times. However, I started planning to start a PhD in IISc and also applied for a Google PhD Fellowship. Somehow luck worked! I received both the fellowship and a PhD position at IISc. Receiving a Google PhD Fellowship made the decision easier, and I returned to IISc to work with the same advisor, Professor Narahari Yadati, I had known during my Master’s. I was also extremely lucky to have Professor Siddharth Barman as a co-supervisor. I also consciously chose to do my PhD in India. I knew going abroad could bring greater visibility and a wider network, but I knew my environment here. I trusted my advisors, had friends and family around me, and felt confident that this was where I could do my best work. That support later mattered more than I realised then.
My research was in fair algorithms for multi-agent decision-making, essentially asking how resources or recommendations could be allocated when multiple people want different things, while still providing mathematical guarantees of fairness. But the hardest problem during my PhD wasn’t necessarily designing algorithms. It was figuring out which problem was worth solving. Fairness in algorithms was still an emerging area. It felt like standing in front of an enormous ocean; there were interesting questions everywhere, yet finding one that was both non-trivial and actually solvable was incredibly difficult.
Early in my PhD, I became obsessed with finding the big problem, the problem that would define my thesis. That pressure took a toll. There were periods when I questioned intensely whether I was doing anything meaningful. Whenever research wasn’t going well, the alternative looked especially attractive: I could have stayed at Xerox Research; I could have been earning well; I could have had a stable job without constantly wondering whether a problem I had spent months thinking about would lead anywhere. That cycle of excitement, doubt, progress, doubt again, happened many times. By the end of my PhD, though, I understood something I wish I had understood at the beginning. A PhD does not have to be one spectacular problem.
The real training is learning how to keep going when you don’t know whether a problem has an answer. It is learning patience, communication and time management. It is learning how to recognise a useful question, learn from the failed attempts, and start again without treating that as failure. I was fortunate to have excellent advisors and a collaborative laboratory. My seniors made the laboratory feel like a community, we discussed research, collaborated, celebrated birthdays and had fun together.
I got married during the second year of my PhD. I had close friends from my Bachelor’s and Master’s in Bangalore, and I lived outside campus. Having people with whom I could talk, complain, laugh and completely disconnect from research gave me a grounding base. I think PhD students sometimes underestimate how important that is.
One of the most transformative experiences of my PhD happened far away from IISc, at the Heidelberg Laureate Forum. Suddenly I was in the same space as Turing Award winners and Fields Medalists, people whose names we knew from foundational textbooks. I assumed students like us would never simply walk up and talk to them. Apparently, they knew that too. At the first dinner, some of these extraordinary scientists went from table to table themselves, talking to students. They spoke not only about their successes, but about mistakes, uncertainty and the difficult periods of their own careers. That week changed something for me. I saw the humility behind some of the biggest names in computer science, and realised that excellence and humility can, and should, go hand in hand.
When you are struggling during a PhD, it is easy to imagine that successful scientists somehow travelled a cleaner path. Meeting people at the very top of their fields and hearing them speak so normally about failure made me realise that uncertainty is not evidence that you don’t belong in research. It is part of research.
That is also why I think prospective PhD students should investigate something far more important than a professor’s publication count- the environment in which those papers were produced. Talk to current students. Talk to graduates. If possible, talk to people who left the laboratory and their experience. Ask what happens when research doesn't go as planned, when students disagree with the advisor, or when a project stops working. An advisor's support can profoundly shape a student's PhD journey. Of course, sometimes issues become apparent only over time; I feel institutions would benefit from a more transparent system that allows students to change advisors when a relationship genuinely isn’t working, without making the process feel adversarial or punitive. And once you begin a PhD, look beyond your own laboratory. Attend workshops, summer schools, conferences, and international forums. Meet researchers outside your immediate field. Some of the most important things you gain during a PhD may come from conversations that never appear in your thesis.
After IISc, I moved to Harvard for my PostDoc. My work also began moving from theory towards more practical problems. I started collaborating with nonprofit organisations in India and Boston, asking how algorithms could help solve problems in maternal healthcare, public health and wildlife conservation. Today, I am an Assistant Professor at Rutgers University. I teach algorithms to more than 200 undergraduate students, mentor young researchers, supervise PhD students, and am building my own research group. It is interesting to be on this side now.
I began as a student trying to understand what research even was. Then I spent years worrying about whether I had found a sufficiently ‘big’ question. Today, part of my job is helping other students find their own questions. When I look back, there wasn’t one dramatic moment when I decided what my life would become. One decision simply led to another.
To summarise, I would advise aspiring researchers not to spend all their time searching for the one problem that will define them. You need to learn how to ask good and relevant questions. Find people who make you a better researcher without making you miserable. Build a life outside your work. And give yourself permission not to know exactly where all of it is leading. I certainly didn’t. My journey happened one question at a time.”
-Arpita Biswas, PhD in Computer Science, Indian Institute of Science, Bangalore
Interviewed and written by Aniruddha Mukherjee