Aakash Indurkhya (BS '16) is a trailblazer in the burgeoning field of artificial intelligence (AI), which continues to spark conversations (and heated debate) in the scientific and political realms. As the co-head of AI at Virtualitics, a Pasadena-based advanced analytics company, Indurkhya is focused on ways to make AI for analytics more effective, ethical, and relevant for real-world use cases.
ENGenuity spoke with Indurkhya to learn about his role within the field of AI and the misconceptions surrounding it, and how Caltech shaped his career path.
ENGenuity: How would you describe what your professional contributions and what you're currently doing?
Aakash Indurkhya: I'm currently the co-head of AI at Virtualitics, which was spun out of Caltech and specializes in intelligent data exploration capabilities. My responsibilities as the co-head of AI include managing different development of AI capabilities baked into our platform and AI solutions that we deliver for our customers using the platform. I joined when the company was only around ten people and now, we are quite a bit bigger than that. Being able to contribute and build this company has been a rewarding experience.
The traditional way that people approach AI is to explore some data set, find a problem you want to solve, start building an AI system like a predictive model or an unsupervised model, and then see if it's possible to get the model to work. The issue with this approach is it leads to a lot of wasted time where you get to an end point and realize the model does not work out so well or the model is not going to address the business need you have. What we do is focus on taking the AI and putting it at the start of the analytic journey—so as you are exploring the data, you are seeing it through the lens of what an AI system would see in the data. This helps you perform a feasibility assessment way faster, and the AI guides you through the exploration. This AI assistance is critical for analysts that have not been formally trained in math or computer science.
ENGenuity: What's the biggest misconception people have about AI and what you do?
Indurkhya: I'm glad you said ‘biggest' because there is more than one. The biggest misconception is that AI is magic and can do almost everything already. Usually, once people are more informed and are trying to figure out how to use artificial intelligence, they start to realize where it makes sense and where it's less relevant.
Also, just because you can build an AI system does not mean you should necessarily use it, because there could be a lot of issues baked into the data that introduce bias or prejudice. If you have bad data, you are basically building in predictions that will give you bad outcomes. I think there are a lot of misconceptions about how mature AI is, which is common anytime you have a technology that has as much hype as this. People tend to overestimate how far along we are because they are focused on the potential. When you get to practicing the discipline, you stay focused on what is feasible right now (and maybe what's around the corner).
ENGenuity: What inspired you to get into computer science and AI?
Indurkhya: My reasons for getting into computer science and AI are different. When I was in high school, I was focused on synthetic biology. I was interested in being able to control genetic circuits and stimulate different control systems within cells. When I got to Caltech, I realized that the parts I found most interesting were building with logic and less about the actual biology angle, although I still find biology fascinating.
What got me into AI was that I had interest in data, even when I was a kid. I've always loved sports and that was probably my first exposure to data. I would open the newspaper, go to the back, and look at the breakdowns of matchups by the numbers. When I was in school, I didn't always have the patience for the math side of things, but with computers, you can automate all of that. I found it interesting to be able to take all those numbers and then do something potentially useful. Around the time I was at Caltech, CS 156 (Learning from Data) was becoming quite popular, and that's how I dipped my toes into the world of machine learning. From there, Caltech has a phenomenal program when it comes to being able to explore things within AI and machine learning.
ENGenuity: What class or professor had the most significant impact on you at Caltech?
Indurkhya: One course was CS 155 [Machine Learning & Data Mining], which was taught by Yisong Yue [Professor of Computing and Mathematical Sciences]. I took that course the first year it existed. To this day, a lot of the fundamentals I've learned in the discipline go back to that course. Another course I enjoyed was BEM 150, which was Data Analytics for Business. That exposed me to the actual use cases and applications of the technology. It was a practical course and there were a lot of projects. Both of those courses informed what I ended up doing with my career.
ENGenuity: How has your Caltech education influenced you?
Indurkhya: There are two main ways that Caltech influenced me. One was being humble and collaborative. Every student at Caltech has the experience of getting there and instantly meeting someone who is 100 times smarter than you ever could think to be, which makes you realize you are going to need help to succeed. That collaboration aspect is celebrated within Caltech. The other thing, and maybe this is specific to the computer science department, was that there was a self-defined track within the computer science major, and you could design your own coursework. I enjoyed taking advantage of that; it allowed me to study things that I may not otherwise have gotten the chance to study. Caltech nurtures your curiosity no matter what direction it takes you in.
ENGenuity: What advice would you give to the next generation of Caltech students?
Indurkhya: As valuable as your Caltech education will be, the thing that will help you as much if not more is your ability to communicate effectively, particularly your ability to communicate very technical things to not so technical people. If you can do that well, you will have an easier time doing whatever it is you set your mind to.
ENGenuity: Is there a project you're most proud of in your career so far?
Indurkhya: I'm really proud of what we are doing at Virtualitics. We've been focused on how to take what data scientists know from studying the craft and reading what is going on in the AI industry and exposing that to people who have not spent their time learning about that stuff and cannot do it on their own. Often, data scientists have a good sense of how to do things, but they don't always know what to do or where to apply their skills. If you can take some of their capability and give it to people that have a keen sense of where they need help and what needs to be solved, then you amplify the power of what one data scientist can do.
ENGenuity: Do you think people are becoming more knowledgeable about AI as it gains more attention from outside of the scientific community?
Indurkhya: With AI at the most zoomed out level, people's awareness started off with things like Terminator and The Matrix. Now, people have slowly been catching on to how AI is showing up in their real life. What you have with ChatGPT, for example, is exposure to a lot more people very suddenly. I think there is more enlightenment, which is encouraging, but there is an equal measure of confusion because they don't understand the steady advancements in the field that led to this new flashy thing. Almost every practitioner I talk to in this industry is not just talking about how exciting it is, but also how cautious we must be in putting it out there and making sure that people do not walk away with the wrong assumptions or understanding of what they have in their hands. Even so, it's been exciting as everyone gets more acquainted and asks more questions.
ENGenuity: What is your favorite story you've encountered recently?
Indurkhya: I've always been a fan of Christopher Nolan's work. When I was at Caltech Interstellar came out and we were lucky enough that Kip Thorne [Richard P. Feynman Professor of Theoretical Physics, Emeritus] gave a talk on the science behind the movie (he served as a scientific consultant for the movie). Being able to watch the movie with all the other students was a real treat at the time! Given that Oppenheimer just came out, that's probably been my favorite movie this year. I didn't know about a lot of the politics that were covered in the movie.
ENGenuity: What is your favorite destination?
Indurkhya: I recently went to Lamu in Kenya, which is a beautiful little beach town just south of Somalia. The culture is unique and has been maintained over a long period of time. There are also donkeys; there are about 6,000 donkeys in a town of less than 30,000 people. That whole area is unique because all the trade routes used to go along the eastern seaboard of Africa and then up and around India and China, so you get a lot of different cultural influences in one spot.
ENGenuity: What gives you the most satisfaction in your work?
Indurkhya: Finding really innovative and well-suited applications of AI and seeing people have that light-bulb moment is really fulfilling. That's when AI stops being a buzzword for them and becomes a real asset to their life. Being a manager and helping people progress their careers and work on things they find interesting is equally important to me. The happiness they take back to their personal life from that may have more impact than anything else I do in a day.
ENGenuity: What keeps you up at night?
Indurkhya: Global warming is a pretty evident threat to humanity. Perhaps I don't literally lose sleep over it, but I certainly appreciate those that are actively combating it. It's really sobering, especially with how fast the ramifications seem to be progressing. As a species we have our work cut out for us.
ENGenuity: What gets you out of bed in the morning?
Indurkhya: Honestly, I'm a big coffee guy and I got an espresso machine that I've been playing with for the past year. I really find the process of making coffee peaceful. When I'm traveling, it's nice to see the variations on it from different cultures as well.