Venkat Padmanabhan
Alums of IIT Delhi
From being a global thought leader in networked and mobile computer systems research to inventing ground breaking inventions like RADAR, persistent connections and pipelining for HTTP, the ProbeGap wireless bandwidth estimation technique and many more, we present to you the journey of Dr. Venkat Padmanabhan in and out throughout his career.
Dr. Venkat Padmanabhan, currently is the Deputy Managing Director at Microsoft Research India. Moreover, he received a Bachelor of Technology Degree in Computer Science and Engineering in the year 1993 from the Indian Institute of Technology Delhi, followed by Master of Science and Doctor of Philosophy from University of California, Berkeley.
Don’t ever feel disheartened, just believe in yourself and just go after what you are interested in. Ultimately, it’s just self-belief ― knowing that ultimately things would just turn out fine. While difficulties in the path can be a little stressful, if you have self-belief then I think you can overcome anything.
When we first came to IIT it was a very overwhelming experience. It is a whole new universe. When we are in school, we are the only toppers and now there is a new culture and seniors rag us. How was your hostel experience?
That is a good question. Many of the best students come to IIT. In general, the best students from various schools come to IIT but my experience was a little different since I was from a very competitive school to begin with. Maybe if you are in Delhi you know of a school named DPS RK Puram. It was a large school then and I’m sure it is even larger now. There were sections where pretty much everyone wanted to take the JEE.
The peer group was already very strong. After the JEE, out of the top 50 ranks that year, 5 were from my section. It wasn’t a huge change for me since 5 of my classmates were in IIT. Back then, we had a hierarchy in our choice of departments and all of us ended up in computer science.
In any system, you select the best and put them together. Everyone is very good but if you start comparing yourself to others, it might put unnecessary pressure on you. In fact, recently I saw this video which said that the students who do the best are the ones who, relative to their surroundings, are doing the best. If you put the 20 best people together then someone out of the 20 will be the 20th. They may be better than the first in many places but since they compete with a strong peer group locally, they may feel that they are at the bottom of the class. I am hoping that people realise that they are in a very select place like IIT so everyone is very good and so something like class rank should not get a lot of attention.
What was the hostel culture like? Which hostel were you a part of?
I was in Kumaon hostel. I was a bit of an outsider to the hostel culture. I knew a bunch of people, but I was not immersed in the hostel culture. My interests were different.
When I entered IIT, I did not get a hostel since there was a shortage of rooms and my family was based in Delhi then. Back then there was the practice of fresh students getting ragged. For the first several weeks I did not go to a hostel, so the good thing was that I escaped ragging almost entirely.
We had a get nice together in 2017/2018 and we went to the hostel, which rekindled old memories.
It is interesting how you first pursued your bachelors from IIT Delhi and then your Master’s in Computer Science. What motivated you to pursue higher studies first instead of starting with a safe job that is usually the common notion even in today’s time?
Things have changed over time, I graduated nearly 30 years ago. Back then, opportunities in India were growing but they were nothing like what they are today, with large corporations, startups etc. Back then, 1991 was when the Indian economy opened up. Before that everything was quite controlled. Then more and more opportunities started coming up. Many people and more in other IITs would look for opportunities abroad as things were viewed to be limited here.
I liked academics and research, so it was a natural path. Before IIT, I wanted to become a physicist, but you do worry about job prospects. So once JEE happened, a bunch of us debated about taking electrical engineering since it is a bit close to physics and then you could pick up the computer stuff if required. But I ended up picking computer science and wanted to do research, so it was obvious to pursue higher studies.
Now also people who want to do higher studies, look abroad. Some people do their PhD in IIT, IISc etc. These numbers are increasing but still most people go abroad. Back then also, most people went abroad. I did my master’s and PhD from UC Berkeley. I had an interest in research and was not thinking of startups. I debated a little between being a professor or an industrial researcher at a research lab and then joined Microsoft Research.
How did you manage your academics at IIT Delhi and did your grades play any role in deciding research as a career option?
The fact that I did quite well at IIT Delhi gave me confidence. I think in the overall institute my CGPA was number 2 but it was also number 2 in my department. So, I didn’t get the silver medal in my department nor the institute medal. My grades were good. I liked the freedom that academics gives you. As a researcher you are the master of your own work. No one can say that these are the 20 things you need to do. Research is a spontaneous thing. You don’t know when a good idea will come to you. I think it’s a combination of interest and my academic record that allowed me to end up at a good university for higher studies.
You had such an amazing CGPA. How did you manage your academics and social life at IIT to create a balance between the two?
As I mentioned, I was a bit of an outsider to the hostel culture, so my social life was not the most active. I think these things are a function of focusing on what’s important for you. It is a question of what your priorities are. I liked academics and was interested in research.
One of the top 5 guys, he decided that after IIT (JEE Rank was 14/15), he wanted to become an IAS officer. He was extremely capable. He just focused on that and didn’t focus on his grades. It just depends on people’s priorities. Some people focused on management and other aspects. In these things, it is best to be clear for yourself what makes you happy instead of looking at others as if it might not be right for you. You might be unhappy, or you might not be the best at it. So just figure out what your priorities are. If you are interested in research and so on, IIT is a great education opportunity for you. These days it is even better, IIT professors were not as active in research as they are now. Now some of the younger faculty and the older as well are very active. As a student you can leverage that as well.
These days the number of people who want to pursue research is significantly decreasing. What is your take on this trend?
Even back then, some of the other IITs had a much higher fraction of students going into research. People from IIT Delhi had a wider set of interests like management, consulting and even startups. Some people opened their own businesses and very successful ones. Very few went for higher studies.
There are two things, the culture of each institute is different. There was a difference between IIT Madras and IIT Delhi as students saw what their seniors did. IIT Delhi had a more diverse set of interests. A good number of people did their IIT degrees and never did engineering later. Many are doing finance and MBA.
Second, the times have changed. Even if you are interested in doing technology, higher studies is one option but you can do a startup or join a big company and grow there. So, there are many more opportunities. You should go with what interests you and drives you. Whether it is product team or research or any other thing, if you go for something that interests you, you will do very well. Students doing research and students doing something else that they are very good at are both just fine.
What would you advise people who are going for research and how should they utilise their opportunities at IIT in terms of academics and network?
It is a completely different world now as compared to then. Back then, we had no connectivity. I didn’t have an email account. In my final year, to correspond with the US universities I was applying to, I had to seek the help of a PhD student in my hostel in atmospheric science. PhD students used to get email accounts. M.Tech and B.Tech did not get one. He was in atmospheric sciences and he had no use for his email account, so he said, “here is my email account and password, do what you want with it!”. So, I would use his account and communicate with universities using it. This underscores how disconnected we were. Today, connectivity and awareness is high. You know what is happening in the world of research. You can look up any paper and any talk anytime you want.
The other thing is that the expectations from students, even B.Techs, have gone up. We all did B.Tech projects back then. My thesis is lying there on my shelf. Today many students publish it. They try to look at something that has got some aspects of novelty and research and publish it. The best students everywhere are doing it. If you are applying for a master’s or PhD program, you are competing with students who as an undergraduate have produced one or more such papers.
I don’t want you to get pressured, but the world is different. You have connectivity. Also, the faculty have become more involved in research. 30 years back, they were also disconnected and did not have access to journals, but now they do. Now especially the younger faculty are very active, so as a student take advantage of the opportunities you have to work with the faculty and also with other researchers through summer internship opportunities at research labs.
Our lab, Microsoft Research India, hosts research interns every summer. Other companies and overseas organisations also offer internships. Use these opportunities to do something interesting. Take the work seriously and try to make something out of it. You do a B.Tech thesis for 6 months to one year so you can try to publish it into an article. Just have greater ambitions than we had back then. We were just happy. When I went for one of my internships, the criterion was not the quality of internships but where I can go, where the work would be light so that I could study for my GRE. If you are interested in academics, use your summers and work the well-regarded people and the good places so you can expand your horizons and impress them so they can act as letter writers for you.
In an increasingly competitive world, it has become difficult for people to pursue higher studies or research as it may not be as rewarding in the beginning as taking up a well-paying job. People have the fear of missing out and they feel compelled to take up a job as everyone else seems to be doing it.
There is no one size fits all here. Everyone is different and everyone’s interests are different. That said, sometimes young people in the IITs have a herd mentality. There is a certain level of comfort with being in a group and going with the flow. I can completely understand that, but it does make sense to not go with the flow and strike your own path. It allows you to follow your own passions instead of what everyone is doing. It gives you more opportunities. And young people certainly have time on their side in pursuing what interests them rather than going with the flow.
There is a herd mentality, not only for jobs or placements, but even as a researcher. In any research field there may be some hot topics like AI/ML is in Computer Science today. It is an area that is important and has exploded. As a student if you are interested in something else like networking or the future of manufacturing, you have less competition. Everyone else may have gone after the hot areas, so there is benefit in striking your own path. With a research career or academic career, you are your own boss. Of course, at an industrial lab, you have layers of management but still you have a lot of freedom to work on what you want and work on your ideas. If that’s the kind of thing you like, I would encourage you to look into research as a career path.
It is not an either-or choice between pursuing research and startups. There are many people who pursue a PhD and therefore are more equipped to do a deep technology startup as they have more technical knowledge under their belt. At UC Berkeley, where I did my PhD, many technology companies took birth. Even a former CEO of Google did his PhD at Berkeley. It is not like you are missing on doing a startup by pursuing a PhD. You can do both. If you are interested in exploring, you should definitely try research.
For computer science and related areas, which is the focus on Microsoft Research, we have set up a research fellow program aimed at bachelor’s and master’s students who are unsure about pursuing research. They can apply for a research fellow position. They are given a two-year position at the research lab and at the end of the two years, some who like research go on to pursue PhDs at good institutes. One of them, very recently he joined the faculty at IIT Delhi, Abhijnan Chakraborty, in computer science. He was a research fellow with me a few years ago. Some of them decide they don’t want to pursue research. I would say some of the smarter ones decide that they wanted to do something else! For instance, the Ola CEO, Bhavish Aggarwal, decided he wanted to do something else. He was one of my research fellows, we used to write papers togethers, I am still writing papers, while he has moved on to higher pursuits! The point is that a program like this gives you the opportunity to see if you like research. This is for computer science and related areas and interested students can consider applying for this. Google, IBM etc also have such programs. It is like an apprentice program of sorts, to help you decide whether you like it do a PhD or do something else.
Also what we have observed is that some people fail to maintain the scale between social skills and research. Some people who are good at research fail to explain to people their research and are not able to express the importance of the topic. How important do you think it is to maintain the scale between the two?
It is a very important point. In fact, I had the opportunity to address the graduating students at the IIT Delhi computer science convocation in 2018. I spoke on this broad topic and one of the things I emphasised on was communication. This was something that we did not emphasise on in my time. Whether it was oral communication or written communication, there weren’t that many opportunities. You would write one BTech thesis. There were no project reports. There were just exams on paper. If it were a project, you would not document the design, goals, alternatives, what worked and what didn’t. We did not have any opportunity to present our work. We did our thesis and our professor signed it. I am sure things have changed now but I think there is no reason that there should be a trade-off between research and communication. For all jobs, communication is important, but for a researcher’s job, it is even more important. In research, you may have the best idea in your mind but if you can’t communicate it, it won’t go far.
Communication means writing clearly and also giving clear talks that grab people’s attention. Sometimes people just make slides packed with information and give a presentation. The audience will likely just go to sleep. Learning how to communicate effectively comes with time. No one taught me this, these are the things that you learn with experience. When I wrote my PhD thesis, I wrote it up and gave a draft to my advisor, he read it and said it was fine. One of the committee members took the trouble of reading it and marking up every other word, not because there was anything grammatically wrong. It was all because of stylistic things and clarity. That was an eye-opening experience for me. It is very important to put yourselves in the shoes of the listener or reader. You should think about what they would want to hear or what they would want to read. I think that really changed me. I became more careful of how I write, and I made sure that anything that I write is clear to someone who doesn’t know a lot about the topic and that the thoughts flow in a meaningful sequence.
So, for anyone in research, if they think that they can be successful without communication skills, they are sadly mistaken. Communication skills don’t require something like a course, it is something you learn with time. These days there is YouTube, where you can listen to good talks by effective communicators. If you are working on a very mathematical topic, it would be a mistake to put all the equations on the slide. Present intuition or a geometrical analogy, which the audience can relate to. If you want to present the mathematical details, talk about the overall picture, and then zoom into the mathematical part. To sum up, I would like to stress that communication is extremely important, for any job and particularly for one in research.
When people think of research as a career the first thing that comes to their mind is a professor and to many people this is not very interesting and right now in this world where even top-notch corporates respect the researchers and get them into the company in the R & D and you have been there sitting among the technicians who create the products, so how has your experience been among the people who have been delivering products continuously and you have been doing research among them?
I will answer the question in two parts. You said academics is not of very much interest to students and I was very much in the same bucket because again partly conditioned by the situation in India where professors are respected and it’s a good career path but it’s not the flashiest or most sought after path. There are many people who go to industry also but if you look the US, where I did my PhD, the top students, or at least many of the top students, actually want to be professors of top universities, so if you want to be a professor at a Berkeley or a Stanford or an MIT, it is very very competitive and very top students compete and that’s because you know universities are the source of lots of ideas, start-ups and next-generation companies. If you look at Silicon Valley a good number of companies are founded by Stanford people and some by Berkeley people and those from other top universities. I just want to make sure that, when they look at academia, students take a broader view. Looked at globally, being a professor doesn’t mean you are just into teaching, which is actually only a small part of being in academia. Idea generation, research and, in many cases, building companies and so on is a big part. This is becoming so in Indian academia too, which has also changed considerably in the past 30 years and will continue to change.
As for working in an industrial lab, when I graduated, I was looking at primarily industrial labs. You would have probably heard of the famed Bell Labs. I had an offer from there and a few other places. Microsoft Research was relatively new then. When I turned down Bell Labs and decided to go at Microsoft Research, some people joked that I had turned down Bell Labs and had instead gone to “Bill Labs”, a play on the name Bill Gates! Anyway, the reason was that, just going back to your question, it is set up in a way that gave one a lot of freedom as a researcher to conjure up with ideas and so on, and at the same time, one is very well connected with the company. So, we are not this ivory tower where we are just coming up with some ideas and publishing it. We are very well connected to the company and so there is an outlet for our work in the company.
I’m not saying every idea of yours would turn into a product, but you know the company has a large enough portfolio of products that you have a good shot at generating interest and finding opportunities for impact with your ideas and that is very satisfying. So, there are multiple things that I have done which either directly got into products or indirectly. An example on the indirect front is the indoor equivalent of GPS that helps you locate yourself in indoor spaces. GPS doesn’t really work inside a mall or airport or something, so I worked over 20 years ago on an “indoor GPS” system called RADAR to do indoor localisation. Microsoft did not capitalise on this by itself, but we did file for patents on this work and those patents were licensed to pretty much all Android handset manufacturers, so essentially my IP was part of what Microsoft was licensing, so there is the satisfaction of knowing that you’re, in some sense, touching potentially billions of users.
To sum up, one of the advantages of being a researcher in an industrial setting is having both direct access to products and the opportunity to impact product through technology transfer. but, again, if someone’s interested in academics, do take a broader view academia. It is not just teaching, and it can be quite enriching and exciting. At least in the US, an academic job is very sought after and I’m sure that will happen in India too.
What made you return to India after working at Microsoft, Redmond for 8 years? What do you think is the difference in the work culture of Microsoft in India and the USA?
Again, it’s a good question. My experience may not be very typical, you know. I was with Microsoft Research in the US for 8 and half years then I moved to India in 2007, so I’ve been here for over 14 years now. Despite this move, I was with the same organisation, Microsoft Research and the way Microsoft Research is set up, it has a basic culture that it is same across all the labs whether its India lab, the China lab, the UK lab or the US labs. So, in that sense my immediate work environment did not really change, since the culture in the lab here is not fundamentally different from the culture that I was immersed in the US lab. That said, there are obviously some differences in the lab environment different countries. I think one the biggest difference is that we have a lot of young talent in our lab here that perhaps was not there in Redmond. For instance, in Microsoft Research India, we have lots of students, who come in as research fellows and are around for 1 or 2 years. These research fellows add a lot of energy to the lab.
On the other hand, even today the centre of action in technology is not India. In many cases, essentially when you look for conferences or if you look at what you study and where the ideas came from, many of them have come from other places, including, significantly, the US, which is a big part of the technology ecosystem. So, when you’re sitting in India you are a bit far away from the centre of the action and so if you have to be aware of that and make sure that you go that extra mile to remain visible at the global level. In terms of going and visiting universities when you go to a conference combine that conference visit with a couple of university visits so that people see you and they get to know you and your work. The research world is a very small and close-knit world, so this network of people is very important. So, when we moved back to India, that was one of the things that myself and other colleagues were mindful of and made sure we were taking the extra steps required to maintain our visibility. But by and large, the work culture for me did not change because I was in the same organization at, just in a different lab.
People tend to change their organisation often after 4 to 6 years but it was not so in your case, so you never had the urge to change the organisation or have different people around you, how did you enjoy your journey?
I think this one is just an individual’s choice. It’s not common for someone to stick around one organisation for a long time. My only little change was that I moved from one Microsoft Research lab to another, but I was still in the same company and organisation. If you’re interested in research or industrial research in computer science, Microsoft Research is an amazing organisation. There are very few such organisations around that are supported by a company that is well funded. Microsoft has been doing well over the years, and it is a company that is willing to invest in research and is willing to give researchers the freedom.
As I said, in our research lab no one ever told me to do this or do that and I also don’t tell anyone, even people whom I manage, to do this or that. So essentially people have the freedom to figure out what they want to do, typically in collaboration with others. Such freedom is very unusual to find elsewhere even in research labs. If you look at some of the other research labs, including the ones that are in India, they are all good, but they don’t have nearly the same level of freedom or support in terms of investment from the parent company. So, in such an environment, there’s really no reason to look elsewhere if you like what you’re doing, because the other options may not provide you the same degree of research freedom. Of course, I’m not saying that Microsoft Research is the only place that provides this. Maybe there are other places too. But let me put it this way — — it is not very common among industrial research labs, whereas if you are in a different kind of job you and are doing products or something else, there are many more options because everyone is shipping products then you can consider company x and company y and so on. However, there aren’t that many places that provide one the privilege of a quasi-academic life, when it comes to publishing ideas, going to conferences, sometimes teaching (I have taught courses at the University of Washington and IIT Hyderabad), advising students etc. but while working in an industrial lab. The old Bell labs was like that, but that was way back. I think Bell Labs has declined in the recent years and decades. So, in some sense, being at Microsoft Research, you are already in a place that is very supportive of what you care about, so there is little reason to move.
How was your experience at the University of Berkeley different from IITD in terms of the research environment and college culture?
I think it was a big change, to be honest, and I said back when I was a student at IITD none of us had an interest in research. I think no one in my class had written a single paper as an undergrad and I think the PhD program was very small. Even among the faculty, I think a very small number of people were actually pursuing research; the focus was mostly on teaching. I mean that’s really what IITs back then focused on. There were exceptions you know, I spent one summer working with Professor M Balakrishnan, who just retired, and Professor Anshul Kumar, both in computer science, and they were doing i think interesting research in their lab, but I think by and large teaching was the focus during my time at IITD.
In comparison, Berkeley was not just very much into research, it was also a completely different world in many ways. Even in terms of teaching, I was in for a bit of a shock. I went there and I became a TA (Teaching Assistant) for a course on computer architecture. In IITD, CS and EE people did take a course on computer architecture. But at least back in our time we had textbooks and we did a small-scale experiment with some 8088 microprocessors, a very basic microprocessor. At Berkeley, the same undergraduate computer architecture course, the students were essentially designing a full-fledged RISC computer as part of the course. And as a teaching assistant, I had to help them, and I had never done any of this before myself. I was a teaching assistant in a position where the students knew more than me!
I’m sure things have changed in IIT since then, but back then, our knowledge was largely from a set of books and theories. There was much less of the hands-on doing the lab work and so on. There was some effort but not nearly as much as is ideal. And part of it was because we used to take 6 courses in a semester, you can’t take 6 courses in a semester and have the time to do hands-on lab work and all. In Berkeley, I think students used to take 2 courses, and then they would do all this lab work. So, from the research point of view, it is a world different but even in terms of classroom instruction and course work I think the hands-on nature of things was a big change for me.
What we have heard from other people is there is a huge difference in what we do in class and the research work in the industry. We are lagging behind a lot, so, how do you think IIT could bridge this gap and students can do something that industries will value at this time?
I would give the answer in two ways. First, I would say that academia should be somewhat detached from the industry. If academia starts working only or primarily on industry relevant problems, they would perhaps be limiting themselves. In academia, you have a lot of freedom to explore things that industry is not even thinking about. Industry may not be interested today but academia could and should be thinking ahead. I think it’s important to be informed about what’s going on in industry and so on, but that doesn’t mean that academics should aim for something that is industrially relevant today because that would be very short-term, and you would not be leveraging your full freedom as an academic.
Second, in terms of engaging better with industry, I think collaboration is a great way. In our own lab at Microsoft Research, we collaborate with many professors on various projects. In computer science at IITD, there is a professor Rijurekha Sen. She was an intern with me when she was a student at IIT Bombay. We used to collaborate with her and her advisor on some projects, and others in my lab had such collaborations too. One of my colleagues is sitting in IITD as an adjunct professor and advising students, while also being a member of Microsoft Research India. I think essentially collaboration, internships, and so on are ways of engaging with industry. But, as I said, you know it’s important to learn industry is up to but don’t limit yourself to just what industry thinks is important today, because then you might be missing out on an opportunity to do something that is much more impactful.
You have had a very inspiring journey and throughout this journey, you might have faced some pitfalls or crossroads. How did you manage to battle through them & what were your major learnings from those situations that you would like to share with us?
I think that is a very important question, certainly, no one’s path is a linear one rising upward. As you said, there are many ups and downs. I can give you some concrete examples. I went to Berkeley — one of the top universities — which was all very good. Usually, you go there and in the first year you take some courses and through those courses, you hope to find an advisor that you can work with.
I took 2 courses in my 1st semester and 2 in my 2nd semester, so a total of 4 professors I got to know and at the end of my 1st year, 3 of those professors retired and 4th one, unfortunately passed away. So, at the end of the first year, there was nobody I could work with. Essentially, I didn’t have an advisor and had to find my own way for funding, which meant a bit uncertainty and stress.
Luckily, I went for an internship at an industry lab -Digital Western Research Lab. Digital used to be a very big company in the 1980s — — it was called Digital Equipment Corporation (DEC). When I went there for the internship, it was the very early days of the World Wide Web. This was 1994, so I think the web, which was born in 91 or 92, was in its infancy and some simple ideas could be tried out to make things better. I worked with this guy, my internship mentor, on this and turned that into my master’s thesis; it had nothing to do with Berkeley. I just did it with this outside guy. So, interestingly, it was only in my 3rd year at Berkeley that I actually found a PhD advisor.
So, out of my 5 years at Berkeley, I didn’t have an advisor in my first 2 years. That was quite stressful because there is uncertainty and you look at other students who have advisors and you don’t have one, but I think what helped me is, of course, I was able to do this internship and the lesson there is also not to put all your eggs in one basket, you know finding professors and all is good but build your network, go for internships, you might meet some great researcher or someone there whom you can learn from and who can act as a mentor.
I think ultimately, it’s just self-belief ― knowing that ultimately things would just turn out fine. While difficulties like not having an advisor can be a little stressful, if you have self-belief then I think you can overcome such things. I’m sure there are other such things also that I cannot recall right now, but even if you look at someone who quite successful, they likely have had ups and downs — — nothing is ever linear. So don’t ever feel disheartened, just believe in yourself and just go after what you are interested in. As I said, everyone has gone through a filter to get into IIT, so everyone is very good and will eventually do well