Senior Machine Learning Engineer | Senior Software Engineer
Senior Machine Learning Engineer with a strong background in AI, deep learning, and software engineering. Experienced in designing and implementing advanced ML systems for image captioning, object detection, and MLOps pipelines. Skilled in AWS services, LLMs, prompt engineering, and serverless architecture. Passionate about continuous R&D on emerging AI technologies including Retrieval-Augmented Generation (RAG).
With a foundation in Computer & Systems Engineering from Zagazig University, I've successfully led multiple projects and worked across various domains including startups, freelancing, and corporate settings. My approach combines technical expertise with business acumen, allowing me to deliver solutions that address real-world challenges.
I specialize in building end-to-end machine learning systems that not only perform well on technical metrics but also solve real business problems. I'm particularly interested in the intersection of computer vision, natural language processing, and cloud infrastructure - creating scalable AI solutions that can be deployed in production environments.
Beyond my professional work, I'm passionate about several areas where AI and technology can make a meaningful impact:
Developed a two-input model for automatic image captioning, combining word embedding and image classification using LSTM and CNN architectures. The system was trained on the COCO dataset and achieved high accuracy in generating descriptive captions for images.
Implemented a system that processes input faces with regression and CNN algorithms to identify 78 key facial points. This project has applications in facial recognition, emotion detection, and augmented reality filters.
Developed a high-accuracy (99%) algorithm that processes 9x9 Sudoku grids in a single shot. Improved upon regular CNN algorithms by consolidating 81 separate outputs into a single output, significantly enhancing efficiency and performance.
Trained a Generative Adversarial Network architecture on a public dataset of actor faces. The system successfully generates realistic human face images with various attributes and expressions, demonstrating the power of generative models.
Created a dataset of 1,000 facial open/closed eye images and implemented transfer learning algorithms to achieve 99% accuracy in the classification stage of cropped eyes. This system has important applications in automotive safety for detecting driver drowsiness.
Led the software team for a wearable exoskeleton robot project that led to a published research paper. The project went through three startup campaigns and was exhibited at multiple venues, showcasing the integration of robotics and assistive technology.
Udacity
Udacity
Coursera (Andrew NG)
Udacity
Coursera
Udacity
Udacity
Huawei
I'm always open to discussing new projects, creative ideas or opportunities to be part of your vision. Feel free to reach out using the form below or through my social media profiles.