Hi, I'm
I'm a 3rd year Computer Engineering student at UBC.
Currently seeking Summer 2027 internships/co-ops (June – August 2027).
My interests are distributed systems, compilers, kernels, and deep learning architectures.
Feel free to contact me if my profile aligns with your interests.
Outside of code you'll find me in the gym or on the basketball court (or playing Maplestory).
Experience
Software Engineer Intern, AI Inference
Tenstorrent / Inference
- → Contributed to a C++ disaggregated LLM inference engine and the tt-mlir compiler stack for PyTorch/JAX
- → Deployed Tenstorrent's first disaggregated speculative decoding setup for Kimi K2.7 with a DFlash draft model on 20 decode and 8 prefill nodes via NVIDIA Dynamo's router, hardening it for its production launch on OpenRouter
- → Implemented engine-side multi-token prediction (MTP) speculative decoding for models like DeepSeek and GLM to speed up token generation, defining an interface with kernel developers for lookahead tokens in MTP prefill
- → Designed a feature enabling prefix caching for linear attention models like Kimi K3, migrating recurrent state snapshots across device memory, host memory, and a global key-value store to reduce TTFT on shared prefixes
- → Developed and extended multi-host distributed runtime systems, using Slurm to orchestrate multi-node jobs and integrating performance profiling tools
Education
University of British Columbia
Bachelor of Applied Science - Computer Engineering
Coursework:
- Data Structures & Algorithms
- Machine Learning & Engineering Applications
- Discrete Math
- Relational Database Systems
- Computer Architecture
- Electric Circuits
- Probability Theory
Douglas College
Engineering Foundations
Coursework:
- Intro to Python
- Object Oriented C++
- Engineering Design
Projects
Real-Time Chat Application
A full-stack real-time chat application with RESTful backend, JWT authentication, WebSocket messaging, and a React frontend, deployed on Render and Vercel.
UBC Course Planner
A cross platform application for planning and organizing UBC courses across semesters with an interactive UI using drag and drop mechanisms.
Lua Bytecode Compiler & Virtual Machine
A bytecode compiler and stack-based virtual machine for a subset of Lua 5.1, built in C++. Implements a full pipeline from source to execution, including a lexer, recursive descent parser, and a VM runtime with closures, lexical scoping, and first-class functions.
Tron Light Cycle Game
A multiplayer Tron light cycle game built from scratch running on a Risc-V cpu softcore on FPGA.
Skills
Languages
Frameworks & Libraries
MPI Tools
MLIR LLVM Linux Git PostgreSQL Docker Kubernetes