Curious by nature.
Practical by training.

I am an Applied Scientist and Machine Learning Engineer. I work across large-scale AI systems, multimodal learning, and ML infrastructure, following research questions all the way to systems people can use.

At Treverse LLC, I design graph-based recommendation models and own their training and serving stack. I am also developing a re-commerce foundation model, beginning with auction value prediction through language-model fine-tuning. Alongside the modeling work, I build shared infrastructure for automated training and deployment.

Before Treverse, I worked as Lead AI Data Science Researcher at Vanderbilt University's AI Negotiation Lab. I led development of an LLM annotation system, reducing estimated per-transcript annotation costs from $5,000 to $3 and increasing human–AI agreement from 30% to 80% in our later evaluation. I also developed negotiator agents and the data and web tools needed to study them. The work was presented at the 2025 AI Negotiation Summit at Harvard and MIT.

Alongside that role, I researched multimodal learning at Vanderbilt's Network and Data Science Lab. I first-authored Inverse-LLaVA, a text-to-vision architecture trained without a separate alignment stage, and released its 7B model weights, reproducible evaluations across nine benchmarks, and analysis artifacts. I also explored graph–language models for temporal interaction data.

I also applied computer vision to archaeological analysis, improving mortar classification accuracy from 60% to 97% with Vision Transformers and reducing visual provenance analysis from months to minutes. Earlier, as an Algorithm Engineer, I delivered vision-based stock detection across approximately 30 warehouses. I owned the pipeline from camera setup and training data through ONNX inference optimization and integration with the warehouse controller coordinating AGVs.

I hold an MSc in Data Science from Vanderbilt University with a 3.98/4.0 GPA and a BSc in Data Science with First Class Honours and Highest Distinction from Beijing Normal-Hong Kong Baptist University.

"Any sufficiently advanced technology is indistinguishable from magic."

Arthur C. Clarke, Clarke's Third Law