As AI companies start to find adoption, we are now starting to see real successes emerge. Arijit Sengupta has built one company and sold it to Salesforce.com. He is doing a second one that is also moving right along. Many insights to draw from this conversation!
Sramana Mitra: Let’s start at the very beginning of your journey. Where are you from? Where were you born, raised, and in what kind of background?
Arijit Sengupta: I was born in India in the Andaman Islands. I spent most of my childhood in Calcutta. I came to the US to study artificial intelligence because at that time even at IIT Kanpur, you could not study AI at the Bachelors level.
I read the book Dimension. The book listed all the researchers who were working on AI. Many of them were at Stanford at that time. I ended up going there.
I figured out that AI is really interesting, but there is more to life. I never really studied Philosophy or anything in the Humanities. Growing up in India, you specialized in science. I ended up getting an Economics degree and missing a Dance minor by one class.
Sramana Mitra: What year was this?
Arijit Sengupta: I was class of 1999, so 1995 to 1999. I went back and checked my passport. I had $240. Of which, I had spent $40 in Singapore. I ended up in the US with $200. I had done a bunch of AI and programming contests in India. I was a geek coming in.
Sramana Mitra: You were exposed to AI at the high-school level?
Arijit Sengupta: Yes, because I just read books. Edward Feigenbaum had written a book called The Fifth Generation.
Sramana Mitra: All I was checking is, there was nothing more than that. Stanford was your first exposure to AI.
Arijit Sengupta: There was a programming contest that they used to have from the science museum community. I was state champion for three to four years. I was national champion a couple of times. Expert systems were very popular at that time.
This was still the time when we were at the end of the expert system hype. I found AI so intriguing because of Asimov’s books to a great extent. I got into so much of robotics because of Asimov. People don’t realize how much of AI is not that complex. You’re basically trying to extrapolate from data. You fall into the same kind of craft in AI that you would fall into any extrapolation from data.
It was a good experience because the professors at Stanford didn’t hype it. Feigenbaum was still at Stanford. In our classes, nobody ever hyped AI. I got a really education out of that.
Sramana Mitra: What happened in 1999? What did you do when you graduated?
Arijit Sengupta: I wanted to go work for Google. Marissa Mayer was my AI TA for one of my classes and she was my debate team captain. I had to get a green card, so I ended up in Oracle instead.
I was at Oracle for almost four years. From there, I went to Harvard Business School. I did a bunch of standards body work. I was on the technical committee for the UNC. I would be in these meetings and I had no idea what EBITDA meant. I didn’t know any of the business metrics.
I went to Harvard Business School to get my MBA. There, I met Clay Christensen who taught this class, “Building a Sustainably Successful Enterprise.” I went to him and said, “Will you be my professor for the Harvard Business Plan contest?” He agreed to be my faculty sponsor and I ended up doing a lot of research which led to my first company called BeyondCore.
Sramana Mitra: What year was that?
Arijit Sengupta: It was officially founded in 2003 but I ended up doing a lot of research on it for a while. Seven years later, Menlo Ventures funded us.