When I started my PhD data analysis a few years ago, I found myself quite alone. Me and my R-Studio – struggling to find answers. Since I did not have anyone on the project who could guide me, or go over my scripts, I was spending excruciatingly long on every problem and making mistakes I wasn’t aware of till much later. In my dire need, I asked my decade-old school friend, Khanjan Damania, who was pursuing his Masters in Business Analytics at the time and has since moved on to work at Amazon, for help. Soon, my learning in R accelerated, and I asked Khanjan to be my co-author. I eventually stopped getting stuck that often, but Khanjan stuck around, always going over my scripts, making sure there wasn’t any oversight. Now that I am at the end of my PhD, in love with coding in R, and all three of my papers in the publication process; I thought of asking Khanjan how his experience had been and his take on academia through his industry-tainted glasses.
Divya – We’ve submitted two of our papers, yaaayy! How was your experience working on academic research papers?
Khanjan – When I started, I didn’t know it was a paper. I knew it was a research, or rather an analysis. Then I registered it was a research, and you had to explain what a research project is. Sometime last year I realised how this all comes down to a paper.
Divya – That is how it started. I asked you to help me with some analysis and then we started collaborating. Did that make a difference for you? From analysis to understanding the research question?
Khanjan – The biggest difference is that you started writing your own scripts and I could see them get better. That was already satisfying for me because I could help you get there. I wasn’t aware of the big picture the entire time. In Industry, we do something, we see the results immediately, we take action on it. The feedback loop in research is longer. So, I feel like I still don’t know the value of what I have added or the value of the research.
Divya – Before our collaboration, did you ever think about being actively involved in research or its publishing process?
Khanjan – When I did my masters, and I had friends doing PhDs, I did think of doing a PhD. But because my Masters was from a Business school, it was all application-based, data collected in real-life scenarios rather than in a lab, and I couldn’t see myself doing research in the projects I was working on. When I started working on Projects in companies, one of my managers was a scientist, and he recommended publishing the work we did. That was the only time I thought of doing it, but we didn’t end up publishing that either. I don’t think it’s for me though, because it seems like such a slow process. All the work I was doing and still am doing, I can see immediate outcomes. I am satisfied with these outcomes, so I don’t feel the need to do more.
Divya – Would you do it again now that you’ve had a taste of it?
Khanjan – Not by myself, but I would definitely help research projects or co-author papers if required. I don’t think I can write a paper by or for myself, mainly because I haven’t been trained in it, so I wouldn’t know how to go about it.
Divya – Would you like to learn the process? If someone told you right now ‘we give you a 3-month break to go learn research skills’ would you do it?
Khanjan – Not really. I like to have a problem at hand. I like helping people. But going so deep into topics doesn’t drive me. I’d rather create generic solutions for problems than pinpoint to something so specific. I’m in no way undermining the work researchers do, but I can’t see myself in it, again probably because I don’t have that training.
Divya – While working on my projects, what differences did you see from your work in the industry?
Khanjan – My first impression when we started a few years back was that your work felt a lot more manual. Industries think of scaling things, so no one has to work on it again. I know we’ve spoken about this topic since, and you’ve told me that more and more researchers are trying to make their research replicable, having GitHub accounts and providing scripts with their analysis. It takes time to get an entire field to start doing these practices, but what you are telling me sounds like it’s going in a good direction. So, the reservations I had then are reducing.
Divya – Yes, unfortunately, I couldn’t do this for my first project, but I am also hoping to do this henceforth. So now if I tell you, in a few years, all the literature is with scripts, reported as detailed as possible. Would that change your impression of science? Why?
Khanjan – Yes, because now there is an overlap between how industry and academia work. I can now read a paper, create dummy data to verify results, and if it works, I believe the article.
Divya – Great! Are there any other differences you still see?
Khanjan – Another point would be, you seem to have more roadblocks to publish. Your work needs to be thorough because crude work will not be published. We can do it iteratively. Our V1 is also good enough. It’s the best we got from nothing. We can get feedback on it, and still come out with V2, V3 and so on.
Divya – Academia can indeed learn this. Like, we could have poster sessions and symposiums only to present what is not done, or what is still underway, to get periodic informal peer review. The last SIPS conference had such poster sessions, and it was terrific. It reduces focus on what is done and achieved and increases focus on improving the work at hand.
Khanjan – Yes, absolutely.
Divya – I came to you because I needed help with my scripts, but did you learn something in the process?
Khanjan – I got exposed to research and science!
Divya – Haha…Ok then, on a final note – What do you think industry could learn from academia?
Khanjan – Research is meant to drive industry, so it’s clear on a conceptual level what industry has to learn. Based on my experiences with you, and based on the fact that I am in a pretty big company, I think we can also learn from you how to solve problems with limited resources. Also, knowledge sharing is standardised and monitored in academia. This is something industry can learn, but I can also see why this won’t be feasible in for-profit organisations.
For me personally, the value addition was ‘how do I think as a subject-matter expert (SME)’. I’m still new in my industry, but I can learn from academia, what level of thinking and what kind of skills I would need to reach the level of an SME. That is what being a scientist teaches you. Trying to overcome problems in an almost perfect way is what would help anyone in industry. A lot of engineering folks want to be that kind of an SME and it takes 8-10 years to reach there. Interacting with researchers will help achieve that and, in the process, they may even publish a few papers. Such a collaboration is a win-win for all.