Selected Projects
/// TECHNICAL_DEEP_DIVES
Metta AI
Open-source multi-agent RL environment studying cooperation and alignment via social dynamics and reward sharing in complex gridworlds.
RL Studio
Open-source RL tooling for designing environments, running rollouts, and launching training jobs on cloud GPUs with real-time analysis and W&B integration.
LensAI
Engineering-focused project involving Swift and BLE for real-time streaming, with on-device TensorFlow Lite inference and embedded C++ firmware plus ONNX acceleration experiments.
OLAF
Open Life Science Analysis Framework for end-to-end bioinformatics workflows, combining an easy-to-use frontend with a Python backend and helper scripts for fast setup.
Quanta
High-performance PDF layout analysis engine using YOLOv8 for document detection and OpenCV for OCR linking. Optimized runtime to 2–5s per page, bridging deep learning with systems engineering.