Where I've worked
and studied.

Columbia University, Department of Electrical Engineering · ELEN E3801 with Prof. Xiaodong Wang.

Tensor & Deep Learning Lab, Columbia University · I implemented L2A, a Transformer-based annealing framework for MaxCut, as a full training and inference pipeline. Batched RL training made each epoch 12× faster, and a CUDA-graph tabu local search cut inference time by up to 10× and GPU memory by 3.5×. On a multi-scale BA/ER benchmark, it matched or beat Gurobi (1-h limit) on all 960 instances with 100–1,200 nodes and outperformed ECO-DQN, S2V-DQN and X2GNN.

IEEE · Special Track Chair for Artificial Intelligence for Scientific Discovery and Creative Art (AI4Science&Art) at the 12th IEEE International Conference on Intelligent Data and Security, and a paper reviewer.

Rensselaer Polytechnic Institute · Tutored more than 150 students in Data Structures, Algorithms and Professional Software Development, one on one, with code review and debugging help in Python and C++.

Rensselaer Polytechnic Institute · I led a 5-member team that formulated quantum circuit synthesis as three MDP representations and evaluated them with Q-Learning and DQN on ten tasks, reaching up to 100% success on Bell-state synthesis and 91% on the Toffoli gate. I then built the first RL benchmark for Ising-model optimization: about 190K instances across 7 problem families, and a Transformer-based RL policy within 1%–9.11% of industrial MIP solvers. The results were published at AAAI and NeurIPS workshops.

Tensor & Deep Learning Lab · I built the participant starter kit (baseline RL solvers and evaluation tools), worked on the contest website and documentation, and reviewed CFP-track submissions.

M.S. in Electrical Engineering

B.S. in Computer Science and Mathematics

Languages

Python · Java · Go · C++ · SQL · TypeScript

Backend

FastAPI · Spring Boot · Node.js · Celery · REST APIs · OAuth2

Data & infrastructure

PostgreSQL · MySQL · Redis · Docker · Linux · Nginx · GitHub Actions (CI/CD) · AWS · GCP

Frontend

React · Next.js · Tailwind CSS

AI & ML

PyTorch · Transformers · CUDA · Reinforcement learning · GNN · RAG · pgvector · FAISS · LLM APIs

Teaching & academic service

Signals and systems · Programming mentorship · Code review · Paper review · Conference organization