Welcome
This is where I write about funny stuff.
Core Expertise & Toolbox
- PyTorch • Reinforcement Learning NumPy • Pandas • MLflow • Weights & Biases • ONNX • ROS2
- Python • C++ Docker • Git
- Graph Algorithms • Dynamic Programming Markov Decision Processes (MDP) • Path Planning
Recent Breakthroughs & Research
⚡ FluxNode | Intelligent Network Orchestration
Developed intelligent resource orchestration for VBS/LTE networks using Reinforcement Learning within the ns-3 simulation framework.
Explore the research on Zenodo📉 Quantization at the Edge | IEEE COMPSAC
Architected a deployment pipeline that achieved a 2.3× model size reduction (from 1.1B parameters down to 470M) for Small Language Models (SLMs) integrated into virtual worlds.
Leveraged TVM, MLC-LLM, and multi-scheme quantization (INT3–INT8) to optimize performance via ONNX and TensorRT (trtexec) directly on hardware-constrained NVIDIA Jetson edge devices.
What Drives Me
I am captivated by systems that close the loop between perception and action. My primary research and development interests lie in:
- Autonomous Driving: Mapping out safe, optimal trajectories through complex, dynamic environments.
- Robotics: Giving physical agents the computational brainpower to reason about their surroundings.
- Edge AI: Squeezing maximum intelligence out of limited hardware, because AI belongs where the action happens.