From Web Dev to Scientific Computing: My Open-Source Journey

I didn't plan any of this. My journey has been less of a straight line and more of a series of "huh, that looks interesting" moments — and following each one further than was strictly reasonable.

It started with the web

My first real projects were web apps. The Happy Attire, an ethnic-fashion e-commerce storefront with a cart, wishlist, and admin dashboard. A smart-campus issue-reporting system on Firebase. A URL shortener, an encryption/decryption app. I interned as a React developer and a Java/web developer, shipping responsive interfaces and wiring up REST APIs.

The web taught me how to ship.

Then the rabbit holes got deeper

Curiosity kept pulling me into harder problems:

• On-device ML → FaceGuard, running neural nets on a phone.
• Edge AI → PredictEdge, squeezing a model onto a Raspberry Pi.
• Scientific computing → SciML, contributing differential-equation solvers in Julia.

Each one felt like a bigger version of the same question: how do you make this actually work under real constraints?

Open source was the throughline

The constant across all of it was open source. Contributing to Meshery (Go, Playwright end-to-end tests, issue triage). Contributing to SciML (Julia solvers and benchmarks). Becoming a GeeksforGeeks Campus Mantri and running workshops to bring others in.

Open source is how I learned the things college doesn't teach: reading unfamiliar codebases, working with maintainers, defending a design decision in a PR thread.

What I believe now

You don't have to pick a lane early. C++, Java, Julia, Go, Python, TypeScript — I learned each because a project needed it, not because of a plan. Breadth, followed honestly, becomes its own kind of depth.

If you're early in your journey: pick the problem that won't leave you alone, and chase it past the point of comfort. That's where the good stuff is.