Experience

Moody's Analytics · Banking team

Building and evaluating AI agents for financial services applications, focused on banking workflows and analyst-facing automation.

CambioML (YC S23) · Full-time

Engineered and productionized AnyParser, a fine-tuned 1B & 2B vision-language model for parsing PDFs into structured Markdown — fully fine-tuned and preference-aligned, beating GPT-4 baselines; SGLang inference delivered 8× throughput on L4 GPUs, deployed as a SaaS on AWS (ECS + Lambda). Led design and deployment of Energent.ai, a Claude-powered Computer-Use Agent (CUA) sandbox with multi-agent orchestration, long-term memory, and per-user Kubernetes-isolated VM sessions scaling to 1000+ users.

Inspur Group · Internship

Built and annotated a custom volleyball dataset and trained YOLOv7 detectors for real-time AI fitness assessment. Implemented YOLO Pose keypoint detection for athletic movement analysis and accelerated inference with TensorRT, reducing latency while preserving accuracy.

Illinois Leadership Center · Part-time

Analyzed weekly leadership-survey data, produced visualizations and reports, and helped facilitate leadership and communication workshops on campus.

UIUC, Prof. Xinzhu Yu · Part-time

Applied deep learning to time-series microscopy of astrocytes for early detection of Alzheimer's-related changes.

National Center for Supercomputing Applications · Part-time

Worked with Prof. Kaiyu Guan and Prof. Sheng Wang on machine learning for geospatial and remote-sensing data, including multi-temporal crop classification on pre-trained foundation models and a tillage-practice detection model spanning Europe.

Illinois Geometry Lab · Prof. Felix Leditzky

Studied quantum capacity bounds via semidefinite programming.