Final-year Computer and Systems Engineering student with strong Python and C/C++ skills and passion for information & probability theories as lenses for learning about the world.
Actively seeking AI Safety opportunities to decrease catastrophic risks from ASI & AGI.
I am an AI engineer with deep passion for AI safety for the last 4 years. Through independent study and coursework I learnt to be able to work with and analyze frontier models.
I am using those skills in developing a product to absorb the second-hand scene in Egypt and the Middle East, I am also working on a verification-acceleration project with Siemens as my graduation project.
Seeking AI Safety opportunities to learn about interpretability & scheming models.
Here are a few technologies I've been working with recently:Developing an ML pipeline to accelerate functional coverage in hardware verification, combining Reinforcement Learning and LLMs to guide SystemVerilog test generation.
Integrating Python-based ML models with a C/SystemVerilog simulation environment via DPI-C, gaining direct experience calling compiled C code from Python and vice versa.
Structuring the project as a clean, modular Python library with well-defined interfaces, argument parsing, and reproducible experiment configuration.
Tech Stack: Python, C/C++, DPI-C, RAGs, LLMs, Reinforcement Learning, SystemVerilog, PyTorch
Building a full-stack application to digitize second-hand stores, with a React frontend and a Python backend integrating the Gemini LLM API, ISBN Detection for automated product tagging and natural language descriptions.
Releasing software intended for real end-users, requiring robust CLI tooling, packaging, and API design.
Tech Stack: Python, React, JavaScript, Gemini API, REST APIs, Supabase, AWS, Docker.
Worked extensively in image processing, computer vision and computational intelligence applications.
Applied multiple deep learning techniques including CNN, LSTM, Attention Models, and Sentiment Analysis.
Implemented Style Transfer algorithms and advanced computer vision methodologies.
Tech Stack: Python, PyTorch, CNN, LSTMs, OpenCV, YOLO
