03
Latest post
Response Engineering for Training-Free Alignment of LLMs
A practical tutorial on Best-of-N, self-consistency, Reflexion, and Mixture-of-Agents for improving model responses without retraining.
Read the notebookAbout
My core expertise is in AI/ML, with a focus on enhancing the capabilities
of LLMs and AI agents. In particular, I apply optimization and machine
learning techniques to improve the quality of data used for both training
and inference, thereby driving better alignment, controllability, and
reasoning performance.
Education
PhD Management Information Systems
Expected Fall 2026
University of Illinois Chicago
MS Business Analytics
May 2021
University of Illinois Chicago
BS Information Systems and Operations Management
May 2020
George Mason University
Research Interests
Preference Learning
Reinforcement Learning
Optimization
Agentic AI
Large Language Models
Machine Learning
News
Follow Son on LinkedIn for recent research updates, presentations, and announcements.
Featured Videos
Training-Free Alignment of LLMs — NAIRR 2026 Tutorial
Hands-on tutorials on aligning language models at inference time without modifying model weights.
Authors: Son The Nguyen, Tommy Cheng, Theja Tulabandhula
Dynamic Tiling for Efficient and Accurate Small-Object Detection
Presented at the UIC IDS Summer Paper Presentation 2023 and SCECR 2023.
Read the paper →Publications
Experience
Graduate Research Assistant
LLM Alignment & Optimization
University of Illinois Chicago Chicago, IL
- Developed MEMETRON, a task-agnostic framework that casts LLM decoding as discrete black-box optimization to find higher-reward responses without retraining or gradient access.
- Engineered CURATRON, an algorithm that recovers missing comparisons and corrects corrupted ones to complete and robustify LLM preference datasets for more reliable RLHF/DPO alignment.
- Developed and delivered a tutorial on training-free LLM alignment as lead instructor at NAIRR 2026.
- Secured a 200,000 core-hour NSF ACCESS compute grant as project lead, powering the lab’s agentic AI and LLM research.
- Advanced a $25K CTRI-funded effort to translate the lab’s LLM alignment and inference-optimization technology (MEMETRON, CURATRON) toward licensing and commercialization.
AI/Machine Learning Intern
AI Avatars
Xpressive.AI Remote
- Accelerated real-time AI avatar generation 4×, cutting latency for smoother, more responsive interaction.
- Improved avatar realism and engagement by refining the underlying machine-learning models.
- Shipped user-facing functionality to create and customize personal AI avatars, boosting personalization and engagement.
Teaching Experience
Lead Instructor — National Artificial Intelligence Research Resource (NAIRR) Annual Meeting
- Training-Free Alignment of Large Language Models: Making AI Safer and Smarter Without Re-Training (2026; with Theja Tulabandhula and Tommy Cheng)
Instructor — University of Illinois Chicago
- IDS 331 — Business Analysis Using Spreadsheets (Spring 2025, Spring 2026)
Graduate Teaching Assistant — University of Illinois Chicago
Graduate level
- IDS 566 — Advanced Text Analytics (Fall 2024)
- IDS 575 — Machine Learning Statistics (Spring and Fall 2024, Fall 2025)
- IDS 576 — Deep Learning and Applications (Fall 2021, Fall 2022, Spring 2023)
- IDS 594 — Machine Learning Deployment (Fall 2021, Fall 2022)
- IDS 561 — Analytics for Big Data (Spring and Fall 2022, Spring 2023)
Undergraduate level
- IDS 200 — Intro to Management Information Systems (Spring 2024)
- IDS 355 — Operations Management (Fall 2023)
- IDS 371 — Business Statistics II (Fall 2020, Spring 2021)
Talks
National Artificial Intelligence Research Resource Annual Meeting (NAIRR 2026)
Northwestern Institute on Complex Systems Job Talk (NICO 2026)
The Workshop on Information Technologies and Systems (WITS 2025)
The Fifth Workshop on Data Science with Human-in-the-Loop (DaSH 2024)
Data Centric AI Workshop (DCAI 2024)
Bi-directionality in Human-AI Collaborative Systems (BHAICS 2024)
Conference on Artificial Intelligence, Machine Learning, and Business Analytics (AIML 2023)
INFORMS Workshop on Data Science (DS 2023)
19th Symposium on Statistical Challenges in Electronic Commerce Research (SCECR 2023)
Midwest Machine Learning Symposium (MMLS 2023)
32nd WORKSHOP ON INFORMATION TECHNOLOGIES AND SYSTEMS (WITS 2022)
INFORMS Annual Meeting (INFORMS 2022)
Database Systems Class at Concordia University of Edmonton, Canada (Guest Lecture 2022)
Awards
University Student Research Challenge
NASA ∙
March 2022
Selected as part of a student team mentored by Bo Zou.
University Doctoral Scholarship
University of Illinois Chicago ∙
August 2021
Graduate College Doctoral Fellowship
University of Illinois Chicago ∙
August 2021
Outstanding Graduate in Information Systems and Operations Management
George Mason University School of Business ∙
May 2020
Excellence in Information Systems and Operations Management
George Mason University School of Business ∙
May 2020
Contact
Email Son at snguye65@uic.edu or connect on
LinkedIn.