Sramana Mitra: What you’re describing is the journey of a lot of techies into entrepreneurship as well. It’s not about technology looking for a problem to solve. It’s about identifying the problem and then figuring out how to solve that problem using technology. In AI, this is a major issue.
Nitesh Chawla: It truly is.
Sramana Mitra: This is why domain knowledge is important. It’s very difficult for AI experts to come up with a problem to solve. They know the technology side of it. They don’t really know what the problems are. We’re seeing this massive interest in the AI segment now for people who have domain expertise.
Nitesh Chawla: Yes, you need that immersion. I am so glad I experienced it. I was a cocky grad student. I quickly realized that you have to immerse yourself in it and figure out what exactly are we trying to solve. I would develop models as well. The model is giving an amazing accuracy improvement today. However, by the time it goes through the vetting and regulatory framework, it could take 10 months to deploy. What’s the point? I can do an incremental change and deploy it within 30 days.
Sramana Mitra: What happens after you finish these two years in banking?
Nitesh Chawla: I got bored a little bit. I needed to do more. Banking is a very safe industry as well, especially Canadian banking. I came to Notre Dame to explore academia. I was gracious that they let me in. I was on a one-year visa. If things didn’t work out, I had to either go back to Toronto, go to India, or find another job in the US in one year. It was a risk worth taking personally, maybe because of my entrepreneurial spirit. I did that in 2005.
A year later, I became a faculty at Notre Dame. I started my 10-year journey. I became an Assistant Professor in 2007. Then I got tenured early. I got tenured in three years. 2011 is when an innovation park was being constructed. I used to look at it and say, “It would be good to have a startup space in there.”
There were certain problem statements that stayed close to me. The world is not going to be solved by just building large data warehouses or large data lakes. I always believed that you need a very specific dataset to solve a specific problem. No one is going to come and say, “I need this business question to be solved.” What used to happen in the bank back in the day was, you’d go to the IT team and get the answer. But my questions will keep coming. How do we make data available that a business user could just use? That was the seed of an idea. That’s when Aunalytics started in 2011. I started discussions with Tracy Graham who’s the Managing Partner of Graham Allen.
Sramana Mitra: What was the beginning of Aunalytics? Who were involved? Did you self-finance? Did you raise money? Who were the first customers?
Nitesh Chawla: When we started, I wanted to stay in academia at the same time. I enjoy teaching and research. Graham Allen provided the seed funding. Then I called to some of my students to join us.
This segment is part 2 in the series : Bootstrapping an Artificial Intelligence Startup with Services: Nitesh Chawla, Founder, Aunalytics
1 2 3 4 5 6