I am Tianyou (Theo) Lai, an MS student in Artificial Intelligence Systems Management at Carnegie Mellon University, originally from Beijing. My research spans model evaluation, reliable LLM reasoning, and human–AI collaboration, and connects them with applied AI and machine learning in healthcare, industrial inspection, metasurface inverse design, forecasting, and energy systems.
In research, I work across the whole pipeline, from dataset preparation and model design to training and experimental evaluation. In industry, I have interned in startup data science, data engineering, quantitative research, consulting analytics, and product data; roles, skills, and service are on the Experience page.
News
- Nov 2026
- Invited to serve as a reviewer for Wiley Expert Systems.
- Nov 2026
- Presented TRACE at ACM MM 2026 in Rio de Janeiro, Brazil.
- Nov 2026
- Serving as a reviewer for The Web Conference (WWW) 2027.
- Oct 2026
- Invited to serve as a reviewer for the IEEE Open Journal of Signal Processing (OJSP).
- Oct 2026
- Invited to serve as a reviewer for ICASSP 2027.
- Aug 2026
- Started the MS in Artificial Intelligence Systems Management at Carnegie Mellon University.
- May 2026
- Served as a reviewer for ACM MM 2026 (Main Track & Dataset Track).
- May 2026
- Presented FREDNet at ICASSP 2026 in Barcelona, Spain.
- Apr 2026
- Our review of federated learning under data heterogeneity is published in Wiley Expert Systems.
- Feb 2026
- Served as a reviewer for IJCNN 2026.
Education
- Carnegie Mellon University — MS in Artificial Intelligence Systems Management
- Lanzhou University — BS in Data Science & Big Data Technology
- The University of Texas at Austin — Exchange Program in Software Engineering
Selected Publications
- TRACE: Training-time Report-guided and Clinically Ordered Concept Editing*. ACM International Conference on Multimedia (ACM MM), Rio de Janeiro, Brazil, 2026. DOI: 10.1145/3767308.3836433
- FREDNet: A Frequency and Decomposed-Spatial Network for Industrial Defect Signal Detection*. IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP), Barcelona, Spain, 2026. DOI: 10.1109/ICASSP55912.2026.11462834
- A Review of Federated Learning under Data Heterogeneity. Expert Systems, 2026. DOI: 10.1111/exsy.70271
- Deep Learning-Based Inverse Design of Broadband Metasurface Polarization Converter. AIP Advances, 2025. DOI: 10.1063/5.0281453
* (co-)first author.
Full author lists and preprints are on the Publications page, and project summaries are on the Projects page.
Interests
- Basketball & snowboarding taking a break from both while I recover from knee surgery.
- Billiards enjoying both the hot streaks and the cold spells.
- Fun facts I have visited nearly every province in China and explored Europe on two two-week trips. I have also completed several high-altitude hikes and climbs above 5,000 m, though a knee injury cost me the chance to attempt a 7,000 m peak. My travel wish list keeps growing.
- 41 / 46 Cities / Airports
- 313h 44m Flight time
- 220,042 Kilometers
- 123 Flights


