Autonomous Driving Technologies will bring about a paradigm shift in the transportation industry by not only enhancing the driving experience but also minimizing human error in road accidents !

Akshita Mittel, our next pathbreaker, Perception Engineer in the self-driving group at Nvidia ( California), works as part of the Autonomous Vehicles team that is responsible for building the self-driving stack for their customers.

Akshita talks to Shyam Krishnamurthy from The Interview Portal about several internships at Amazon, Google and Facebook Reality Labs, and varied projects in computer vision, that shaped her career path in Deep Learning technologies !

For students, if you want to pursue a career in AI, try to focus on real world applications that yield societal benefits !

Akshita, can you share your background for our young readers?

I was born in Lucknow, Uttar Pradesh. My father is in the Merchant Navy and holds his Engineering Degree from D.M.E.T (now known as the Indian Maritime University). My mother did her PhD from Banasthali Vidyapith in Education and Psychology.

Growing up, I was never in one city for a long time, as my dad’s job had us moving from coast to coast. While shuttling around had its challenges, I consider myself fortunate that I had the opportunity to see such a diverse set of cultures. 

As far as my personal background, I had an innate fascination for arts and craft and was naturally good in mathematics, which I guess shaped most of my life.

My parents always strived to expose me to a varied set of experiences growing up, so I was always swamped in extra-curricular activities, such as tennis, football, swimming, Girl Scouts, KUMON, to name a few. I think this instilled a mentality of always wanting to explore no matter what, which continues with me till date. I am hoping it will continue to do so. Curiosity not only instills a zeal for life but also gives me an opportunity to talk to many interesting people.

What did you do for graduation/post graduation?

I completed my undergrad from IIT Hyderabad in Computer Science and Engineering, and went on directly to pursue my Masters in Computer Vision at Carnegie Mellon University.

What were some of the drivers and influences that led you to a career in Computer Vision?

To be honest, my father was the key driver of my career at the beginning, as it is for many children. Growing up, I knew just one thing, that I wanted Math to be a key major. Attending KUMON (After School Math & Reading Programs) religiously, I always pushed my boundaries when it came to Math and that used to excite me. In fact, I always had my eye on Oxford’s Math department for my graduation. It was my mother who introduced me to  the world of computers. She was always fascinated with computers having studied DOS in the late 80s and continued with Microsoft’s certification courses throughout my childhood. Together they taught me Math, Physics, and Computers (my Chemistry was never at par).

Later during my days at IIT, I slowly fell in love with Machine Learning, especially the applications of Computer Vision (using AI on videos and images). Honestly, it wasn’t until my final year of undergrad when I realized that I had subconsciously selected most of my key projects within the field of Computer Vision. That’s when I decided to pursue my Masters. Once again, I found out what I liked simply by doing.

How did you plan the steps to get into the career you wanted? Or how did you make a transition to a new career? Tell us about your career path

I joined IIT Hyderabad back in 2013. In 2014, Google came up with a STEM program to mentor girls early on in their career in order to keep them more adhered to careers in tech. Once again, I was fortunate enough to be selected as a part of that program. I had an amazing internship under Vamsi Krishna Katakam and Prathima Devarasetty. Together, they patiently introduced me to Google’s monorepo which fascinated me at the time. Based on my project, I was called back for a second internship in my third year. 

Here I worked with Sourav Dasgupta on Google Docs sharing policies. I remember looking up to him as a brilliant coder during that internship. I did do something interesting during my second internship at Google though. I took the time out to schedule 1:1s with many of the engineers, senior engineers, my previous mentors, and managers. You see, if you ask, people are generally very accommodating when sharing knowledge. I asked about the different projects going on at Google, picked the brains of VPs about what growth would look like, what new projects I could look forward to and so on. 

At the same time, I decided to take some of the tutorials and workshops that Google had in Machine Learning. This is when I realized that I wanted to step away from pure software engineering and work towards my masters. By the midpoint of my internship, I had started preparing for my GRE. Back at IIT, I had a host of amazing professors, but I would like to mention three of my professors. First off Prof. Vineeth N Balasubramanian, my Bachelor Thesis advisor (Machine Learning and Vision). Prof. N.R.Aravind, my algorithms professor: I did many small research internships with him on algorithms and graph theory, which helped develop an intuition for deep thinking. Finally, Prof. Ramakrishna Upadrasta, my Compilers professor, pushed me in all aspects of my life. Together, they helped with my grad school applications right from selecting colleges to building my portfolio. They really guided me so that I could land CMU.

At CMU, I got the opportunity to work at Oculus (Facebook Reality Labs) under the guidance of Prof. Simon Lucey, and later went on to intern at Amazon Lab126. During these internships, I got a chance to work on varied projects in computer vision. I worked on a small piece of Oculus’ code to make their 3D avatars more real (photorealistic). At Amazon, I worked on making indoor home security cameras more smart.

How did you get your first break?

I think working on the latest tech at these companies is what eventually got me selected to work at Nvidia’s Self-driving team. 

What were some of the challenges you faced? How did you address them?

I think my trajectory has been smoother than a lot of my colleagues. I look at them and genuinely feel inspired to simply work harder. Having said that, I do have some lessons that I learnt.

Lesson 1: Be yourself as authentically as possible. People will always have views on you. Life isn’t fair, and at different points in life, different people will get different opportunities. Make the best of what you have and don’t be shy about it. Celebrate your achievements and celebrate the achievements of those around you louder. If you work hard towards a goal, life will find a way to give it to you, even if it’s not straightforward. Unfortunately, at different stages in life, I have faced backlash for my achievements and background as well. I always kept this in my mind. So, when I see someone doing better than me, I cheer them on.

Lesson 2: Always be curious and work hard. Never fear the struggles. There are times when everything in life seems hard. For instance, missing out on a promotion or not getting the job you wanted. I remember the pandemic being a difficult time for all of us, not just emotionally but in terms of job security. I remember not cracking IIT in my 1st attempt and watching all my friends go to college living their lives in their new campuses. In moments like these, all I think about is just putting in the work and the rest will fall into place. That has seemed to work for me so far. Another advice a friend once gave me was, stop doing things because you must or because it is in your coursework. Explore things you are curious about. Best advice ever.

Lesson 3: Take risks and don’t be afraid to stand your ground. The two main risks I took in my life are the two things that have paid off for me splendidly. I chose to take the drop/gap year for IIT because I believed in myself, plain and simple. Later, I chose Nvidia over Google when going for my first job, simply because I liked the project and team that I was going to work with at Nvidia. Money does matter, but some things matter more. At the time, Google was paying significantly more, but somehow I didn’t feel I would be happy, so I took a risk, and now it is paying off.

Where do you work now? What problems do you solve?

I work as a perception engineer in the self-driving group at Nvidia (Autonomous Vehicles). While Nvidia is famous for making its hallmark GPUs, they really are involved in everything pertaining to AI. They have a fantastic Autonomous Vehicles team which works on building the self-driving stack for their customers. Within that, I work on writing the software to make the car understand what obstacles are on the road when driving, what speeds they could be going at, to name a few things.

What skills are needed for your role? How did you acquire the skills? What does your day to day look like? What do you love about your job?

I think the main starting skills required are coding, machine learning, and computer vision, and the knowledge on how to write decent software. While I learnt most of this during my courses, projects, and internships, a lot of the information is now available on courses on Udemy and Coursera for instance. All you have to do is search for free courses to see if you have an interest in them.  My day to day is pretty typical. I wake up, exercise. Either work from home or catch a train. Then I work on various projects, come home, talk to loved ones and in the remaining time explore hobbies and latest advancements in tech. I am passionate about understanding and helping other people understand the latest tech.

The part that I love about this job is actually the way people think. There’s a lot to learn from the latest research which I appreciate, but when I see people come up with simple yet brilliant solutions that actually drive products to completion, I tend to learn a lot. 

How does your work benefit society? 

I think that the whole self-driving industry is a paradigm shift as far as the standard of living is concerned. With just the introduction of automatic cruise control, lane keep, and blindspot detectors, driving has transformed for many people. When we bring self-driving solutions into play, it not only revolutionizes the truck industry but the mode of transportation itself.

Tell us an example of a specific memorable work you did that is very close to you!

I can’t get into the specifics, but at Nvidia I had the opportunity to work on a product end to end. Right from demonstrating the proof of concept, to implementing the training code, to actually see it being used in the final product. Now I work part time on that project to maintain it. The sense of ownership was special.

Your advice to students based on your experience?

I think I covered this in detail when I described the challenges and lessons I learnt in life. Be authentic, don’t be shy, be curious, there’s no shortcut for hard work, and be willing to take risks.

Future Plans?

For now I still have things to learn and explore at Nvidia. In parallel, I plan to be more involved in efforts that help bring knowledge on tech advancements to everyone. Honestly, my career has just started, and I am excited to see what the future holds.