Proof-of-work portfolio

Hi, I'm Brian Hsu

Computer Science & Cognitive Science student at the University of Michigan.

I build high-performance systems — from HPC clusters and scientific computing platforms to data pipelines and ML research tooling — with an eye for how software supports real human and scientific workflows.

Proof paths

Infrastructure Systems

HPC clusters, bare-metal provisioning, distributed systems, networking

Data & Analytics Systems

Pipelines, measurement, operational analysis, data-informed decisions

Research / ML Tools

Scientific computing, ML workflows, cognitive science tools

Selected work

systems

HPC Cluster Build

From-scratch bare-metal cluster: provisioning, InfiniBand networking, and verification

Rocky Linux · iPXE · SLURM · InfiniBand / RDMA

systems

Distributed Cache Layer

High-throughput in-memory caching for microservices

Go · gRPC · Consistent Hashing · Docker

research

Cognitive Load Tracker

Bridging HCI research and real-time software instrumentation

TypeScript · React · Python · Flask

work

ETA Prediction Pipeline

Internship project at Lyft — maps & routing

Python · Spark · Airflow · Feature Stores

Experience

Supercomputing Engineer

Los Alamos National Laboratory

Building HPC clusters from scratch with Rocky Linux, iPXE, and SLURM — InfiniBand/Ethernet networking and benchmarking 400Gb/s+ data transfers.

Scientific Computing Engineer

Memorial Sloan Kettering Cancer Center

Shipped an Open OnDemand dashboard for MSK's HPC cluster, cutting new-user support tickets by 20%. Built real-time SLURM monitoring and automated usage reporting.

Data Analytics Intern

Lyft

Built SQL pipelines analyzing 44k+ users; proposed scoring changes that reduced operational costs by 7% and negative user costs by 12%.

Skills

Python · C/C++ · Linux · AWS · Docker · SLURM · React · PyTorch

B.S. Computer Science & B.S. Cognitive Science · University of Michigan, Ann Arbor

Contact

Download Resume