Electrical Engineering Graduate · C++ · Machine Learning · Digital Design
I build projects to understand how systems work—from implementing neural networks in C++ to designing a processor in Verilog.
My work spans software, machine learning, and hardware, with a growing focus on real-time computer vision and embedded systems.
Built a CNN in C++17 with explicit forward and backward passes, modular classes, batch normalization, and dropout.
98.55% MNIST test accuracy · C++17 · Eigen · CMake
Trained both architectures from scratch across four image-classification datasets, exploring how their performance changes with task complexity.
Python · PyTorch · Computer Vision
Implemented attention, positional encoding, and encoder–decoder blocks to explore the inner workings of the Transformer architecture.
Python · PyTorch · NLP
Developing a modular C++ vision system, starting with camera capture and object detection, with eventual deployment on Raspberry Pi.
VisionCapturePipeline: frame capture, processing, display, and recording.
ObjectDetector: image preprocessing, model inference, and detection postprocessing.
| Area | Technologies |
|---|---|
| Programming | C++, Python, C, Assembly |
| Machine Learning & Vision | PyTorch, OpenCV, CNNs, Transformers |
| Systems & Tools | Linux, Git, CMake, TCP/IP |
| Hardware | Verilog, digital design, signal processing |
| Embedded & Robotics | Raspberry Pi, ROS |
- MIPS 32-bit Microprocessor — processor implementation in Verilog.
- TCP Noisy Channel — communication-system simulation using C and TCP/IP.
- CNN Architecture Guide — CNN implementation and architecture exploration in PyTorch.
- AI for Robot — robotics and SLAM experimentation using ROS and Linux.