Hello there

I build a lot of things.

Some of them are systems projects—GPU work, compilers, low-level tooling, the kind of stuff where you end up caring about memory layouts, execution, and why something is slower than it should be. Some of them are just things I wanted to exist, so I built them.

A lot of my earlier work came out of UMass (research, iCons, STORMM, tra86), but I’ve also been building personal projects alongside that—things like a native bitmap editor for macOS, an Apple Watch app for lightsaber hit detection, and an LLM-based automation system that turns natural language into actual scripts.

There isn’t a single theme across all of it, but the common thread is that I like understanding how something works and then building something practical on top of that. Sometimes that ends up being low-level systems work, sometimes it’s a small app that solves a very specific problem I had.

Current things I'm working on

• Splatr — a simple, native bitmap editor for macOS
• SelfSelect / Automator — turning natural language into executable automation
• SaberStat — Apple Watch IMU-based hit detection (still tuning signal processing)
• STORMM (under PsiThera) — contributed to build systems, GPU/CPU compilation pipeline, and packaging for a large-scale molecular mechanics codebase; reduced build memory from ~70GB to ~19GB and improved CI times significantly.

I also write on my blog, Declarative

Why sponsor?

This page is just a way to support the process.

If you’ve found anything I’ve built useful (or even just interesting), sponsoring helps me:

• spend more time building instead of context-switching
• clean up and open source more of my work
• actually write things up instead of leaving them half-finished

No paywalls or gated content, just more freedom and consistent output.

What you get

• Occasional devlogs / writeups
• Early looks at what I’m building
• A say in what I work on next (within reason)

Links

• 🌐 https://skushagra.com
• 💻 https://github.com/suobset

Some projects are serious. Some are questionable. Most of them taught me something.