cv

Bachelor of Science

Carnegie Mellon University

Mathematics and Computer Science

• Machine Learning
• Algorithm Design
• Computer Systems
• Software Engineering
• Theoretical Computer Science
• Probability
• Algebra
• Operations Research

Research Assistant

RoBot Intelligence Group, CMU Robotics Institute

Conducting research on world models for soft robots under Prof. Jean Oh and Prof. Jeffrey Ichnowski.

• Collected multiview datasets of soft-finger deformations in 2000+ states.
• Reconstructed 3D time-series models using VGGT and neural predictions.

Research Assistant

DROP Lab, CMU Robotics Institute

Worked on Video-Generation Environment Representation (VGER) for 3D reconstruction and motion policy generation.

• Integrated video generation with 3D foundation models for single-image reconstruction.
• Developed a multi-scale noise framework for motion-consistent environment representation.
• Second-author paper published at NeurIPS 2025.

Research Assistant

Data Interaction Group, CMU Human-Computer Interaction Institute

Developed machine learning systems for healthcare applications under Prof. Adam Perer.

• Co-authored CHI 2025 paper 'Tempo' for collaborative predictive model design.
• Built ML components (Dense, LSTM, Transformer) and integrated SHAP-based interpretability.
• First-authored ML4H 2025 paper 'TempoQL'—a human-readable query system for EHR data.

Research Assistant

WiSE Lab, CMU Computer Science Department

Working on OpenFLAME: a large-scale federated localization and mapping system integrating indoor and outdoor 3D maps, 3D tiles, and search infrastructure using multimodal LLMs.

• Built an interactive 3D map platform with CesiumJS and Google 3D Tiles.
• Developed a Dockerized 3D tile service with Django for map delivery.
• Implemented a multimodal LLM-based search engine for indoor object discovery.

AI Intern

Soterea Automotive Technology Co., Ltd.

Worked on deep learning–based driver assistance systems.

• Collected and processed real-world driving data.
• Trained and tested danger-detection models to improve driver safety.

DOSE3: Diffusion-based Unified Out-of-Distribution Detection on SE(3) Trajectories

under review

Introducing Students to Research and Reproducibility with Open Science Tools

ASEE Annual Conference 2024

Tempo: Helping Data Scientists and Domain Experts Collaboratively Specify Predictive Modeling Tasks

CHI 2025

OpenFLAME: A Federated Spatial Naming Infrastructure

arXiv:2411.04271v2 (2025)

From Single Images to Motion Policies via Video-Generation Environment Representations

NeurIPS 2025

TempoQL: A Readable, Precise, and Portable Query System for Electronic Health Record Data

Machine Learning for Healthcare (ML4H) 2025

Distributed Bitcoin Miner

Implemented a distributed Bitcoin mining system using a custom reliable UDP protocol.

• Implemented LSP protocol in Go for fault-tolerant communication.
• Built job scheduler with fairness and fault recovery.

Cupid Dating App

Developed an AI-powered dating app with chatbot and matching system.

Memory Allocator and Proxy Server Projects

Built systems-level projects including a 64-bit memory allocator and concurrent proxy server with LRU caching.