Electrical Engineering graduate with a strong technical foundation in software engineering, machine learning, computer vision, and embedded systems.
I enjoy building projects that combine software, algorithms, and real-world systems — from implementing neural networks from scratch to developing C++ vision pipelines for embedded and robotics applications.
- Programming: C++, Python, C, Assembly
- Machine Learning: PyTorch, CNNs, Transformers, Vision Transformers
- Computer Vision: OpenCV, image classification, real-time vision pipelines
- Systems: Linux, Git, TCP/IP, CMake
- Embedded & Robotics: Raspberry Pi, ROS, edge AI
- Hardware: Verilog, digital design, signal processing
A complete CNN implementation built in modern C++ using object-oriented design.
Implemented convolution, pooling, fully connected layers, activation functions, backpropagation, optimization, dropout, and batch normalization.
C++ · Deep Learning · OOP
A practical comparison between a Convolutional Neural Network and a Vision Transformer for image classification.
Explores architecture design, training behavior, regularization, optimization, and model performance.
Python · PyTorch · Computer Vision · Transformers
Implementation of the Transformer architecture with attention, positional encoding, encoder-decoder blocks, and training pipeline.
Python · PyTorch · NLP
I'm currently expanding toward real-time computer vision, embedded AI, and robotics.
My current development path is:
Camera → Vision Pipeline → Object Detection → Edge Inference → Robotics / Action Planning
with a focus on deploying efficient systems on devices such as the Raspberry Pi.
- CNN Architecture Guide — practical CNN implementation and architecture exploration in PyTorch
- TCP Noisy Channel — communication-system simulation using C and TCP/IP
- MIPS 32-bit Microprocessor — processor implementation in Verilog
- AI for Robot — robotics and SLAM experimentation using ROS and Linux
I like building systems from the inside out — understanding how they work, implementing the core ideas, and then applying them to practical problems.