Projects
Fun stuff built with 🤍

Personal Website
A Next.js, Tailwind CSS project which is based on Next App directory with React Server Component and uses Contentlayer to manage markdown content. It is my go-to place for sharing recent discoveries and experiences, as well as testing out front end technologies.
Read more →
Distributed Key-Value Store
I built a fault-tolerant, linearizable key-value store in Golang using the Paxos consensus protocol and sharding for scalable load balancing. The system leverages goroutines, channels, and gRPC over TCP/IP to handle concurrent requests in a multi-threaded environment while tolerating network partitions and node failures.
Read more →
Operating System Kernel
I developed a full-featured OS kernel in modern C++20, with support for preemptive multitasking across 50+ CPUs using custom spinlocks, mutexes, and condition variables. The system includes a Unix-style virtual memory subsystem with swap-backed and file-backed pages, copy-on-write process forking, and a network file server with crash consistency guarantees.
Read more →
Out-of-Order RISC-V Processor
I architected a superscalar, out-of-order RISC-V processor in SystemVerilog based on a MIPS-R10k pipeline. The design supports speculative execution, early branch resolution, byte-level store-to-load forwarding, and multi-banked caches with customizable associativity. I also automated testing and regression workflows using shell scripts for scalable verification.
Read more →
Embedded Systems and Vision
I developed a full embedded stack integrating STM32 and Raspberry Pi using UART, I2C, and GPIO peripherals. I implemented real-time sensor data acquisition, stepper motor control, and a dual-camera OpenCV pipeline for Rubik’s Cube color detection, followed by a two-phase solver for autonomous solving.
Read more →
Internship & Co-op Scraper Platform
A full-stack platform for scraping and managing internship/co-op listings. Built with Next.js and Supabase, it integrates Google OAuth for login, uses Python (Beautiful Soup + asyncio) for multi-source scraping (LinkedIn, PittCSC, Ouckah), and supports saving, filtering, and tracking applications.
Read more →