Mohamed Hamdi

• Developed GPS-Based Property Search: Designed and implemented a GPS-enabled property search system, allowing users to efficiently locate and explore properties based on specific criteria.
• Optimized Media Content Services: Enhanced video pre-rendering using the FFmpeg library and improved image pre-processing through tools like Pillow, OpenCV, and machine learning techniques to optimize media content delivery.
• Refactored & Optimized Legacy Codebases: Improved functionality, performance, and maintainability of legacy projects through systematic code refactoring and debugging.
• Exclusive Streaming Platform Development: Played a key role in developing and optimizing a Netflix-style exclusive streaming platform for The Fellas, a well-known British YouTube channel, which successfully reached 700K users within the first three months.
• Lead Backend Developer for AI-Driven Social App: Led backend development for a mobile application that enables users to generate and upload AI-created avatars and match with others, similar to Tinder. Designed and implemented secure, scalable backend infrastructure to support real-time interactions and media processing.
• API Development & Documentation: Designed and implemented scalable, secure web APIs while authoring comprehensive API documentation to facilitate seamless communication with front-end teams.
• Test-Driven Development (TDD): Developed unit tests following TDD principles to ensure software reliability, robustness, and maintainability throughout the development lifecycle.
• Deployed applications using containerization technologies such as Docker, ensuring consistent runtime environments across development, testing, and production stages.

• Built a Python-based web scraping bot using Selenium/Playwright to collect a diverse dataset of food images for model training and evaluation.
• Trained and evaluated a transfer learning model on the curated dataset, applying advanced machine learning techniques to improve classification accuracy. using TensorFlow.
• Deployed the trained model as a web microservice using Flask and Docker, ensuring scalability, reliability, and ease of deployment.
• Developed a React-Native mobile application that allows users to upload images and utilize the trained model for real-time food recognition and classification.

Source Code: Hama101/PFE_FLASK_API (GitHub), Hama101/PFE_MOBILE (GitHub)

Developed a text-based chatbot using GPT-2, leveraging natural language processing (NLP) techniques to emulate the speech patterns and mannerisms of movie characters, such as Harry Potter.

• Processed and curated character-specific text data to enhance linguistic consistency and authenticity in chatbot responses.
• Implemented and optimized model training, validation, and deployment workflows to ensure efficient learning and response generation.
• Deployed the chatbot as a functional chatbot on discord, exploring the creative potential of NLP in character emulation.