Hello, Welcome to my den.

I am an MSc research student in Computer Science at Western University, working with Dr. Mike Domaratzki on how machine learning models can better represent future environmental conditions for crop performance prediction. Before starting graduate research, I completed my BSc in Computer Science and Engineering at BUET, and worked as a software engineer at Chaldal.

Research Interests

I am currently enjoying my research work at Western in Bioinformatics and AI. It has given me room to experiment with modern machine learning models and think about agricultural prediction from a different angle. At the same time, I have gradually found myself becoming more interested in programming languages, database systems, distributed systems, systems for machine learning, GPU programming and kernel optimization, and low-level systems research.

Software Engineering

My software engineering experience after graduation was short — just 6 months. But it was enough to make me fall in love with the craft of building software. Working in a production environment with a new programming paradigm taught me to appreciate how complex real-world codebases can be and how much discipline it takes to ship reliable features. It was both exciting and humbling.

It also helped me internalize an important lesson: every line of code is a liability for future maintainers. That’s why it is worth spending significant time on modeling the system code. The decisions we make today are often paid for much later. I would rather make that cost as light as possible.

Since then, even while busy with research, I have kept building Projects on the side because I enjoy the process of learning by doing. I tend to build tools that would solve my own daily problems, or projects that would force me to implement systems primitives and core protocols from scratch. It is my way of peeling back the layers of the tools we use every day.

Competitive Programming

Competitive programming is where I learned how to think. Over the years, I’ve solved more than 1,600 problems across Codeforces (where I reached Specialist), AtCoder, CSES, and USACO. I also spent many weekends competing in contests like Google Hash Code, Kick Start, and the ICPC Dhaka regionals.

It did more than teach me algorithms, it trained me to be obsessive about constraints, edge cases, and implementation details. Most importantly, it made me comfortable sitting with a hard problem for hours, learning how to find structure in the noise.

Work Ethic

I believe that almost everything I learn finds a way to become useful later, even when it seems unrelated at the time. I have experienced it too many times to ignore it. We live in a world where more and more things are being abstracted away. My goal is to look under every abstraction possible in my work, step by step. Whether I am working with a model or a service, I want to understand what is actually happening.

I value clarity, correctness, and maintainability. I do not believe in “Clean Code” as dogma, but I do believe in clear business logic. Simplicity is what I try to achieve in my work, although it rarely comes easily. I appreciate the value of collaboration, but I also have a deep sense of ownership over what I build.