SoCurious Spotlight: Joyjit Chowdhury, Senior Data Science Engineer

Joyjit Chowdhury is a data engineer turned data science practitioner and machine learning engineer with an infinite appetite for learning and solving problems (specifically with various AI/ML/DS toolkits).

In his current role as a senior data science engineer at Socure, he works with an amazing team of data scientists and ML engineers on research and development efforts for some, “cool and interesting AI-based products.” He can be quite modest.

How the Endless Journey of Learning Keeps His Wheels Turning

Joyjit is extremely passionate about extracting and discovering data, exploring interesting patterns, and experimenting with ways to create data models at scale to serve real-time inference pipelines. This work of his means that Socure customers get the most accurate results possible.

But nothing comes even close to his true passion: learning. It’s an endless journey that keeps him motivated, engaged, and passionate about his work.

Born and educated in India, now living in Phoenix, he’s been an engineer professionally for 15 years and yet his appetite for learning continues to grow. He regularly takes courses to keep his knowledge and skills sharp by taking advantage of providers like Coursera, Edex, AWS, IBM, Microsoft, Linkedin, Springboard, Databricks and others.

A Deep Appreciation for Mathematics

So where does this deep desire for learning come from? Well, linear algebra, for one. Joyjit often remembers staring outside the window during linear algebra classes in college or blindly crunching EigenVector/Value calculations in numerical analysis with some forsaken programming language. As he explains, linear algebra is such an awesome thing, I still get amazed with its applications all around us, on the hidden side of all the smart UI’s we take for granted every day.”

It’s part of a deeper desire to understand how people think, why they do what they do, and how our actions impact everything around us. He finds that mathematics (and data) makes these things easier to grasp, despite not always being able to finitely define them, and he brings this to his work at Socure which is why his teammates enjoy working with him so much.

The brain, for example, began to intrigue and mystify him as he continued to learn and develop his engineering knowledge. It was only after wandering into the endless ever beckoning world of machine learning that he truly realized what an amazing computing and storage unit our brain really is.

For all that the brain seems like a miracle, Joyjit feels we take its magical powers for granted; even trivial activities involve extraordinary processing, like driving a car. But he says the real power of this little machine becomes clearer if we attempt to simulate its actions artificially, like with self-driving cars.

Coding a self-driving car involves training huge and complex models with tons of visual and audio data, along with other driving-related behaviors. Yet as human beings we can simply drive with ease, making split-second assessments on a myriad of factors in real-time (and sometimes while speaking on the phone). No one even thinks twice about it.

Inquiry Happens in Unlikely Places

Beyond pondering how autonomous vehicles learn compared to how the human brain does, he even contemplates learning models during conversations with his child’s preschool teacher.

At a recent teacher meeting, the discussion (driven by Joyjit) drifted to approaches of how numbers and arithmetic operations can be taught based on the basic theory of counting the actual value instead of memorizing an operation with a number. A positive integer and an operation on two such integers should be something that the kid should be able to intuitively define with something countable.

This made Joyjit wonder how analogous his own learning of matrices and determinants in high school and college were based entirely on a computational perspective-memorizing formula, and not understanding the geometrical intuitions of such operations. It was when he was studying principal component analysis when he went back to EigenVectors and how it determined PC’s of higher dimensional data in a lower dimension, and then from there, he went back all the way to determinants and matrices to realize what they do in linear algebra geometrically. This realization blew his mind. He explains “All that time and I never knew all these things clearly, I was just writing code to spin up 10 digit precisions.”

And all that from a conversation with his son’s preschool teacher.

Inspiration From An Unlikely Childhood Hero

Joyjit’s childhood hero was the genius mathematician Srinivasa Ramanujan, who was a mostly self-taught prodigy who changed the world of mathematics forever. The fact that he had so many breakthroughs in such a short life keeps Joyjit hungry to keep learning and exploring the world of mathematics.

Joyjit also loves to help others learn, which is why he started his Medium blog. His favorite quote is from Ulysses, by Lord Tennyson, and should come as no surprise: “To follow knowledge like a sinking star, beyond utmost bound of human thought…”

Want to work with people like Joyjit? Check out our open jobs here.