hi, i am Saptarshi.
This is mostly a small archive of things I have studied, built, read, abandoned, returned to and thought about longer than I probably needed to.
I study at IISER Bhopal. Most days are some mix of machine learning, data, maths, research, too many open tabs and wondering where the week went.
so far
A few things I remember more clearly than the rest.
I spent most of my school years at Don Bosco in West Bengal. I liked computers, books, puzzles and things that gave me something to figure out.
Most of it was ordinary in the best way. School, home, books, computers and years that felt slow until they were suddenly over.
I picked science for senior secondary school and almost immediately the world shut down.
Classes moved online, sleep schedules stopped making much sense and JEE, JEE Advanced, IAT, WBJEE and a few other exams occupied a fairly unreasonable amount of my life.
Not my favourite era. Strange how far away it feels now.
I joined IISER Bhopal and eventually found myself spending most of my time around data science, machine learning, mathematics and research.
University has mostly been a mixture of interesting ideas, unfinished side projects, questionable sleep and new subjects I suddenly decide I need to learn. Somewhere in between, the years started moving a little faster.
I wanted to see how far I could get making a little game builder with plain HTML, CSS and JavaScript.
It eventually became a browser based editor with tiles, physics, logic and the ability to export games as standalone HTML files.
Basically, a tiny Mario Maker that got slightly out of hand.
There is also a 3D version. It is still a bit unstable and occasionally behaves like it has personal issues.
This was one of my first proper research experiences. I worked remotely with Professor Imon Mukherjee and got a fairly early look at how messy research can actually be.
Ideas changed, experiments did not always work and the paper we hoped for never really happened. Still learned a lot.
I worked remotely with Professor Robin Evans and spent some time learning about Markov models, Bayesian networks and graphical models.
It was one of those experiences where every answer somehow created three more questions.
My first offline research internship.
I worked around EEG signals, biometrics and machine learning, while also getting much better at figuring things out when nothing works on the first attempt.
I worked with John Thomas on removing artifacts from scalp EEG signals using deep learning.
It ended up being a nice mixture of neuroscience, signal processing, machine learning and many hours spent looking at waveforms.
My first proper corporate internship.
I worked around cybersecurity, NIST frameworks, security processes and risk assessment.
Very different from academic research, which was probably exactly why it was useful.
away from the laptop, sometimes
Research is thankfully not my entire personality.