Projects

Things I've built and worked on.

10 Jan, 2025

A minimal implementation of SGLang for understanding LLM inference optimization techniques including continuous batching and KV cache management.

llm-inferencepythonresearch

PythonPyTorchCUDA

18 Dec, 2024

Diagnostic tool for analyzing vLLM request latency, KV cache utilization, and queue behavior under load.

toolingpythonperformance

PythonPrometheusGrafana

1 Dec, 2024

Interactive visualization tool for exploring attention patterns in transformer models.

visualizationtypescriptresearch

TypeScriptReactD3.js

15 Nov, 2024

Library for tracking and debugging GPU memory allocations in PyTorch applications.

toolingpythoncuda

PythonCUDAPyTorch

1 Nov, 2024

This very website - built with Next.js, TypeScript, and Tailwind CSS. Features blog posts, project showcase, and responsive design.

webtypescriptopen-source

Next.jsTypeScriptTailwind CSSshadcn/ui