How did Perplexity AI Founder Aravind Srinivas study AI in IITM despite being an Electrical Engineering Student?

During his time at IIT Madras, Aravind, an EE student, discovered the world of machine learning through friends in computer science and an encouraging guide, Professor Ravindran, who championed cross-departmental learning. 

‍

“I learned about deep learning, RL, all these things before. Now it's obviously the hot thing. At that time it was not, and IIT Madras was probably the only campus at that time that even had a professor working in AI and machine learning.  So that clearly made a big difference in my life. If I fell somewhere else, I would not have been doing what I'm doing right now.” said  Mr. Aravind Srinivas, the founder of Perplexity AI.

‍

Aravind in his first year failed to switch majors to computer science. However, this setback turned out to be a blessing in disguise, as his background in electrical engineering provided a strong foundation in the essential concepts for machine learning. He credits a Python programming class taught by a forward-thinking professor for equipping him with the Python-centric world of ML.  “Nobody was really into Python that much in IIT at the time, and he was very prescient, and that helped me a lot, because obviously, ML is largely being done in Python,” said Aravind. He fondly recalls late-night study sessions with friends at IIT Madras, bonding over assignments and contemplating the future during leisurely walks. These moments of camaraderie and shared academic pursuits shaped his experience at the institution.

This early exposure to deep learning and reinforcement learning, at a time when AI was still in its nascent stages, laid the foundation for Aravind's future endeavors. During his time at IIT Madras, Aravind published in top tier AI conferences like International Conference on Learning Representations (ICLR), AAAI Conference on Artificial Intelligence, and NeurIPS. Aravind collaborated with Prof. Ravindran B and Prof. Mitesh Khapra at IIT Madras, and had a total of 9 papers published. Aravind later went to University of California Berkeley for his PhD to work on problems in reinforcement learning, generative models and representation learning. After graduating from Berkeley in 2021, Aravind worked at OpenAI on language and diffusion text models. In 2022, Aravind co-founded Perplexity AI which recently joined the unicorn club. More here.