I'm a CS and mathematics student at Georgia Tech working on world models, computer vision, and ML systems. I previously worked on production ML data pipelines at Meta. I like projects where model behavior and software reliability both matter.
LinkedIn · Explore my repositories
- Robot learning: hierarchical world models, goal-conditioned policies, and planning in simulation.
- ML systems: reproducible training and evaluation, data validation, and model monitoring.
- Software engineering: reliable backend services, concurrency, APIs, and full-stack tools.
Compare learned high-level dynamics with worker-driven rollouts for goal-conditioned navigation. Shared baselines, matched evaluation seeds, and explicit model comparisons.
Python · PyTorch · Offline RL · Simulation
A webhook delivery service with idempotent enqueue, leased workers, crash recovery, retries, and dead-letter replay. Includes a React dashboard, PostgreSQL integration tests, and CI. A portfolio implementation with at-least-once delivery, not a claim of production deployment.
Go · PostgreSQL · React · TypeScript · Docker
Local deepfake-risk monitoring for video calls, with face detection, score smoothing, configurable thresholds, and an Electron interface.
Python · Computer vision · Electron
TurtleBot3 navigation in Gazebo with SLAM, AMCL localization, path planning, and a Python goal-sequencing node. Simulation only.
ROS · Gazebo · Python
GPT-2 fine-tuning experiments comparing quantization profiles, cyclic precision training, and robustness to typo attacks under a fixed training budget.
PyTorch · Hugging Face · LoRA · Model evaluation
A reversible-change workflow for coding assistants: split risky changes, plan migrations, review diffs, and define rollback paths before merging.
Developer tools · Testing · Code review
ML and robotics: Python, PyTorch, OpenCV, Hugging Face, NumPy, pandas, scikit-learn, ROS/Gazebo
Backend and full stack: Go, SQL, PostgreSQL, TypeScript, JavaScript, React, Node.js, FastAPI
Engineering: Git, Linux, Docker, CI, unit and integration tests
I'm interested in software engineering and ML engineering roles where I can build, test, and understand the whole system.