01 / Index

To build intelligent
systems & make
technology human.

Hi, I'm Thang (Brian). I am

Scroll to explore >
Brian Truong

Thang Thoi Truong

Software Engineer · AI Researcher · PhD Student

Four years at UT Austin sharpened my craft across full-stack engineering, data visualization, and NLP research. I completed my honors thesis under Professor Mishra on multilingual language models, and shipped real-world AI tools through UT Austin's Liberal Arts Technology team.

I'm now working on my PhD at Auburn University, advised by Professor Anh Totti Nguyen ↗︎. My research sits at the intersection of explainable AI, visual intelligence, and human-centered systems — building models that not only perform well, but can show why they do. My research PageGuide ↗︎ puts that philosophy into practice.

Outside the lab, you'll find me at the cinema ↗︎, on the tennis court, grinding Leetcode, or at the gym.

Selected Work

Built &
shipped.

Research systems, AI tools, and full-stack products I've designed, built, and shipped.

News

Aug 2025 Started PhD at Auburn University. 🐯
May 2025 Graduated from UT Austin with honors. 🎓
May 2025 Published chapter on Cardiovascular Risks — CRC Press. 📄
Apr 2025 PageGuide paper published on arXiv — browser navigation extension. 🔬
02 / Work

Experience

2024 – 2025
NLP iSchool Research Group

Natural Language Processing iSchool Research Group

NLP Developer · UT Austin, TX

  • Architected a multilingual NLP pipeline translating 89.6 GB of parallel corpora into 28 languages using SeamlessM4T
  • Co-authored PolyverseVLM — a lightweight vision-language model using 80M SigLIP and 500M Qwen2
  • Optimized batch parallelism for model fine-tuning, reducing training time by 40%
PythonBashSeamlessM4TQwen2
2022 – 2025
Liberal Arts Instructional Technology Services

Liberal Arts Instructional Technology Services

Junior Software Developer · UT Austin, TX

  • Developed 20+ full-stack web apps and dashboards with modular codebases and REST API integration
  • Migrated legacy Drupal systems to Laravel, implementing advanced filtering and user management
  • Revamped three WordPress websites, increasing user engagement by 40%
  • Implemented PDF viewer for UT Instapoll, supporting 50,000+ users across UT Austin classrooms
PythonLaravelNode.jsDockerRTableau
2023 – 2024
AI Health Lab

AI Health Lab ↗︎

NLP Research Assistant · UT Austin, TX

  • Applied BERTopic with UMAP + HDBSCAN to uncover 20 thematic clusters across 12,146 LLM paper abstracts
  • Validated clustering integrity using 5 comparative techniques including PCA and dendrogram analysis
  • Surveyed 12 foundational LLM papers, identifying 3 previously unreported research themes
BERTopicPythonUMAPMatplotlib
Summer 2023
Claudius Legal Intelligence

Claudius Legal Intelligence Inc.

Software Engineering Intern · Remote

  • Developed 4 production-ready Django web pages with interactive frontends from Figma designs
  • Led UI/UX testing with legal professionals, improving usability scores by 18% in internal testing
PythonDjangoFigmaBootstrap
03 / Research

Publications

02
arXiv 2025

PageGuide: Browser extension to assist users in navigating a webpage and locating information

arXiv preprint · arXiv:2604.23772 · 2025

T. Nguyen, T. T. Truong, R. Zhou, T. Bui, C. Agarwal, & A. T. Nguyen. (2025). PageGuide: Browser extension to assist users in navigating a webpage and locating information. arXiv preprint arXiv:2604.23772. Read Paper ↗︎
01
CRC Press 2025

Integrating advanced sensing technologies and artificial intelligence for predicting cardiovascular risks

CRC Press · Cutting-Edge Diagnostic Technologies in Cardiovascular Diseases · pp. 180–219

Truong, T. T., Truong, T. T., Tran, B. N. V., Sabet, C. J., Rutledge-Jukes, H., Odat, R. M., Luu, D., Le, M. H. N., Nguyen, T. T. H., Theologou, T., Harky, A., Nguyen, D., & Huynh, P. K. (2025). Integrating advanced sensing technologies and artificial intelligence for predicting cardiovascular risks: A data-driven approach to modern healthcare. Read Chapter ↗︎
04 / Play

A Bit More About Me

What I work on

Explainable AI &
Visual Intelligence

I build models that don't just predict — they explain. My current work at Auburn focuses on making AI systems more transparent, interpretable, and human-friendly. If a model can't show its reasoning, it can't earn trust.

Recent work: PageGuide ↗︎ — helping users navigate complex webpages using AI.

When I'm not coding

Cinema. Tennis.
Weights. Repeat.

I'm an AMC A-List member and have watched 40+ films in the last year — yes, I track every one with a rating. On weekends you'll find me on a tennis court or in the gym.

Currently obsessed with: my movie log ↗︎

How I think about AI

Amplify, don't
replace.

The best AI is a multiplier. I believe in building systems where the human stays in the loop — curious, empowered, and still the one making the final call.

Output = You + Max(AI) not Max(You, AI)

Leave Your Mark

Draw anything — a doodle, an emoji, a signature. This canvas is yours.

05 / Contact

Get in Touch

Have a question or want to collaborate? Drop me a message.