-
I like compilers, machine learning, programming languages, math and other stuff; I also dabble a fair bit in fullstack development. Broadly, I like software abstractions that make life easier for both humans & hardware.
-
π¨π½βπ» I currently work on machine learning compilers at Qualcomm. My work includes optimizing PyTorch lowering techniques and inference compilers within the AI Cloud Suite.
-
π¨π½βπ I studied CS & Math at Cornell University, with a concentration in Programming Languages (PL) and Compilers. While there, I worked on a couple of fun things:
-
π§± Led the development of Cornell's course management software at CMSX with Dr. Andrew Myers.
-
π¬ Researched compilers, programming languages and hardware for networking and packet scheduling at CAPRA with Dr. Adrian Sampson and Anshuman Mohan.
-
π¨π½βπ« Put [fun] in functional programming at Cornell (FA23-SP25). Raised critters at Cornell's Honors DS & OOP course (FA25). Taught people who know much more than I do how to build great software at Cornell (SP26).
-
I'm fond of music, birds, linguistics, baking and sports of all forms as well.
For a more in-depth look at what I do, visit my website kabirsamsi.com. Feel free to connect with me over LinkedIn.
- C++, Python, Rust, C, Java, TypeScript, OCaml, Scala, Kotlin, Haskell, Swift, Julia, Ruby, HTML/CSS
- Machine Learning: PyTorch, TensorFlow/Keras, Pandas, Hugging Face
- Compilers/Low-Level: LLVM, MLIR, CUDA, Triton, Assembly (RISC-V, x86)
- Fullstack: MERN, SQL, Angular, Vue, AJAX, Flask
- Other: Rocq Prover, Kubernetes, Docker, LaTeX, Git
Check some of my larger public projects on GitHub including Saberchat (platform for schools & community organizations) and FindYourBird (ID and information site for amateur birders). At some point I'll get them back on the internet. In my spare time, I like to hack on:




