Hello, I’m Nguyen Phuc Nguyen#

I am a Ph.D. candidate in Systems Engineering at Boston University with a background in Computer Science and Physics. I work on problems that sit at the intersection of software systems, optimization, and applied machine learning, where careful reasoning about state, scalability, and failure modes matters as much as raw performance.

Alongside my academic research, I build independent software projects to explore system design, tooling, and end-to-end implementation in realistic settings.

Education#

Ph.D. in Systems Engineering | Boston University (Expected 2026)
B.S. in Computer Science & B.A. in Physics | Syracuse University (2021)

Research & Professional Experience#

I am currently a Graduate Research Assistant at the Paschalidis NOC Lab, where I apply machine learning and systems methods to clinical and theoretical problems.

Production ML Pipelines
Designed, built, and deployed an end-to-end ML pipeline trained on a 100,000-patient dataset, now running weekly inference in support of a clinical trial on hypertension prescription at Boston Medical Center.
I also engineered a data pipeline processing a 10-year dataset (30,000 appointments across 6,000 patients) to predict missed CT screenings and enable targeted patient support.
Algorithm Design
Developed a spectral algorithm for discovering latent policies from demonstrations, guaranteeing global convergence in a single data pass and avoiding common failure modes of Expectation–Maximization methods.
Latent Process Modeling
Contributed to the design and implementation of a flexible framework for discovering latent processes, with applications in bioinformatics (mutation analysis) and remote sensing (hyperspectral unmixing).

Selected Projects#

Mokuro Library #

A self-hosted, multi-user library for reading and editing OCR-backed comics and documents, designed to run on home servers or NAS environments.

This project grew out of a desire to have a reader that treats OCR data as first-class, editable content rather than a static artifact, while remaining practical to self-host and maintain.

Highlights:

Server-backed library and storage
All content and metadata live on the server filesystem, avoiding browser storage limits and enabling access across devices and users.
Multi-user design with private state
Users share a common library while maintaining independent reading progress, bookmarks, and OCR edits.
Interactive OCR reader and editor
Documents are displayed with selectable OCR text overlays, and OCR output can be corrected directly in the browser by editing text and adjusting bounding boxes in place.
Non-destructive OCR editing
OCR changes are tracked as structured edits rather than overwriting files, enabling undo/redo, reset to the original version, and reconciliation with upstream updates.
Performance-aware document viewing
Long documents are rendered using virtualized scrolling to avoid performance and memory issues when viewing hundreds of pages.
Reliable export for large archives
Volumes and collections can be exported as ZIP or PDF files using streaming downloads, avoiding large in-memory buffers and supporting stable downloads for large libraries.
Designed for self-hosting
Runs locally via Docker Compose with minimal configuration, making it suitable for trusted LAN or VPN setups.

Matter Phase Simulation #

A systems-oriented simulation project exploring emergent physical phenomena using Rust.

This project implements a particle simulation with an efficient grid-based method to visualize non-ideal gas behavior, crystalline formation, and annealing processes, using the Bevy engine for visualization.

Quantum Virtual Machine (QVIM)#

An experimental project exploring language design and metaprogramming in Julia.

QVIM provides a domain-specific language for expressing quantum logic circuits with readable syntax, along with a simulator that optimizes gate operations via basis changes.

Technical Skills#

Domain	Tools & Languages
Languages	Python, TypeScript, Julia, Rust, Bash, SQL, C++, C, Lua
ML & Data	PyTorch, scikit-learn, Hugging Face, pandas, Weights & Biases, cvxpy, Jupyter
Systems & Dev	Docker, Git, Linux/Unix, Nix, Cargo, Conda, LaTeX
Core Competencies	Data Structures, Algorithms, Probability, Statistics, Optimization, Linear Programming

Connect#

GitHub: https://github.com/nguyenston
LinkedIn: https://linkedin.com/in/nguyenston
Resume: View PDF