We’d like to introduce you to Vishwas Mruthyunjaya, Senior Data Scientist at Megagon Labs. We’ll discuss his growth at Megagon, his advice to aspiring data scientists and engineers, and his interesting journey from Robotics to AI.
What led you to pursue computer science?
I was first introduced to programming during my elementary school years. It was during computer lab hour when we were taught the basics of DOS programming. One of the main incentives for doing well in the lab was the opportunity to play the game Dangerous Dave, which motivated me to finish the programming tasks quickly and better (maybe?).
As I progressed through high school, I had the opportunity to assemble my very first computer with the help of a friend. The process of building something from scratch has always been exciting to me, and this experience was no exception. However, I must admit that my focus at the time was primarily on gaming, and I would occasionally tinker with DLL files in an attempt to hack my favorite games. As I spent more time gaming, I began to develop a deeper appreciation for computers and their inner workings, particularly in the realm of game theory–spoiler alert! I did not pursue game theory; instead, I got my degrees in computer science and robotics.
While I initially had aspirations of pursuing a career in mathematics and physics, my experiences with computers during my elementary and high school years played a significant role in shaping my future trajectory. Additionally, the typical expectations from my Indian parents to pursue medicine or computer science also influenced my decision to focus on computer science.
How was your educational journey into AI?
During my undergraduate studies, we were primarily focused on software engineering fundamentals, the software development lifecycle, and learning programming languages like C, C++, and C#. Unfortunately, there was no AI course offered at my school at the time. However, a select few of us were interested in pursuing AI and convinced a new professor to teach an elective course. This professor had to teach himself AI to put together the curriculum and teach the class. For the course to be approved, we had to recruit at least 15 students. We successfully convinced enough students to register and became the pioneers of AI courses at my university.
After completing my undergraduate studies, I landed an internship at Infosys and then eventually was hired as a systems engineer. During my time there, I worked extensively with databases and state machines. Looking back, I believe that my early hands-on work with state machines helped pave the way for my interest in natural language processing (NLP). Specifically, I was dealing with a lot of text data and found myself looking for ways to automate manual tasks. It was this drive to streamline the work that ultimately led me to explore the world of NLP and expert systems.
After two and a half years as a systems engineer, I felt that pursuing a master’s degree would be the next logical step in my career. As someone who has always loved soccer, I decided to explore opportunities abroad in England–an incentive being the ability to watch soccer matches locally. The University of Plymouth quickly caught my attention, as it offered a path to Carnegie Mellon University (CMU)–a highly esteemed institution that I simply couldn’t pass up. I went on to complete two master’s degrees in robotics, one at Plymouth and the other at CMU, with a focus on human-robot interaction.
Isaac Asimov, the acclaimed author of the Foundation and Robot series, was a major inspiration for me to pursue robotics. My passion for building interactive systems that require an understanding of the world led me to focus on human-robot interaction, which necessitates the use of AI. I am incredibly grateful to Carnegie Mellon University for providing me with the opportunity to pursue my master’s degree there, as it truly shaped my future.
What interests outside of work do you have or practice?
I love pursuing my many interests, and one of my favorite hobbies is building things. I love building Legos, and I’m particularly drawn to space-oriented sets like NASA spaceships, Star Wars, and space stations. I also have a passion for physics, and I enjoy stargazing, doing some astrophotography (even though I’m not very good at it), and reading popular science books on the subject.
In addition to these bookish pursuits, I also enjoy some outdoor activities like playing soccer and solo hiking. But there are times when I like to sit down and build robots from scratch, like a Lego robot or Arduino rover. I’ve even been looking into buying all the parts for a Mars rover that you can build at home. Apart from my “building” hobbies, I go to music school to learn to play guitar.
In addition to pursuing my hobbies and career, I strongly believe that education is the cornerstone to solving any issues socially or personally. That’s why I am always eager to collaborate with mentors and friends–to learn and contribute to the transformation of education. I am thrilled to be participating in an upcoming education and AI workshop with CMU, as I believe that utilizing technology to create a more effective education system is crucial. Providing people with the tools and resources they need to succeed is essential, and I am committed to doing my part in contributing to this effort.
How did you go from robotics to NLP/KGs/ML with Megagon Labs?
My background in human-robot interaction has led me to focus on bringing together robotics and natural language processing (NLP). As more robots begin to roam around, it’s important to ensure they have good conversational capabilities so they can properly respond to commands. This is why I’m particularly interested in language; it’s essential for human-robot interaction.
While I previously worked at a robotics company, I wanted to shift my focus to language interaction. This led me to work for a conversational AI company, where I gained valuable experience in NLP. While I miss working on physical robots, I believe that NLP is central to the development of future robots. I’m constantly seeking opportunities to gain more experience in NLP so that I can continue to use these skills in the future to build better robots. During my time at both, I gained a deep understanding of the different layers of conversation and was able to build language models from scratch. This experience showed me how vital language models are for many NLP and speech recognition tasks. Since then, I’ve been focusing on using language modeling tools and language models, either fine-tuning or building from scratch, to become a specialist in the field.
The transition to working at Megagon was a significant change for me, but I found myself intrigued by the company’s focus on knowledge graphs (KGs) in some of the projects. While I don’t have much background in KGs, I see it as a challenge to learn more about them while continuing to grow within the field of NLP. As a research lab, Megagon presents a fantastic opportunity to expand my knowledge and skills, and I was excited to take advantage of it.
I had a great experience working at the conversational AI company, but I found myself missing the research aspect of the field. That’s why I was excited to be a part of Megagon, where the team is knowledgeable and supportive. They are down-to-earth and easy to work with, and there’s a good balance between work and personal life. I was particularly drawn to Megagon because of Estevam and Eser’s strong backgrounds in research and industry. After working for two startups, I’m grateful for the emphasis on work-life balance here. It was an easy choice to make.
How has working with the team helped you grow so far?
The eight months of me working at Megagon has been a great learning experience. Nikita and Pouya have been fantastic mentors, especially in the area of KGs where I don’t have a strong background. We have regular discussion sessions where we talk about experiments and problem sets, and I’ve been able to learn a lot with their support. The weekly reading group is also highly helpful in understanding what’s going on in the field currently and seeing different perspectives from within Megagon. Collaborative dialogues during these sessions have been impactful, and I appreciate the opportunity to learn from my colleagues.
The one-on-one sessions with Eser, Estevam, and Nikita have given me a high-level understanding of the direction of Megagon. Their experience and feedback have helped me learn how to do certain things differently and make corrections when necessary. Although I know that I still have much to learn, this kind of immediate feedback has been immensely helpful in my learning and growth, and I’m grateful for the emphasis on mentorship and collaboration at Megagon. It’s a refreshing change from my past industry work.
What advice do you have for those in the industry?
It’s important to continuously challenge yourself and explore new opportunities, even if you’re satisfied with your current job. Staying up-to-date on industry developments and gaining hands-on experience through personal projects can be a great way to expand your skills and knowledge. It’s also important to not rely solely on your company or school to provide you with everything you need to succeed. Taking chances on yourself and seeking out new opportunities, whether that be through online groups or personal projects, can lead to discoveries and growth. Remember, there’s no one set way to learn and grow, so find what works best for you.