Data Science student, Faculty of Artificial Intelligence and Informatics, Horus University of Egypt (Damietta) — graduating 2028. I work across the stack that connects AI to the real world: telecom data science day to day, digital twins whenever I get to choose. Mission: help make AI in Egypt world-class.
Currently:
- 🎓 Microsoft Student Ambassador
- 👨🏫 Data Science Instructor & Lead @ MLSC (Microsoft Student Club, Horus University)
- 🗂️ PMO and Tech Lead @ HUNA
Previously: Data Trainer @ Outlier AI · Data Analysis Summer Intern @ Orange Egypt
Foundations (self-taught, ~7 years in) — HTML/CSS → Python, OOP via Pygame, design patterns, backend, automation, early IoT.
Exploration — Flutter app development, hardware + IoT integration, DevOps, and enough Linux breakage to pick up cybersecurity basics the hard way.
AI & Data (current) — production-level ML connected to real backends, cloud deployment (Huawei), computer vision, agentic AI/LLMs, a little data engineering — and digital twins, which stuck as the favorite.
🔧 USSD Bulk Dialer — Flutter/Android client project. Layered architecture: Riverpod, go_router, sqflite, Kotlin Foreground Service. Releases →
🎮 Corruption: The Journey — original Metroidvania-style action-platformer (Pygame), Hollow Knight-inspired, with a corruption-force combat mechanic and branching outcomes. Repo →
- AtmoSmart — IoT air-quality monitor with an agentic AI summarizer and a voice assistant for blind users (repo · live)
- AtmoTwin — environmental sensor digital twin: FastAPI + InfluxDB + Grafana, physics-driven simulation, anomaly detection, what-if scenarios
- Raqeeb — IoT child/item-tracking Flutter app; BLE + NB-IoT + LoRa + mesh, agentic AI on top (repo)
- AI CFO — agentic finance assistant (FastAPI, PostgreSQL, Redis, Celery, Next.js)
- Iris Classifier — end-to-end ML with FastAPI + web frontend + tests, ~97% CV accuracy (repo)
- Factory Defect Detector — PatchCore + ResNet18 + FAISS anomaly detection on MVTec AD
- DeepFalcon (ML4SCI GSoC) — Conv-VAE and GNN classifier on jet-image data (repo)
- Ma'at — CV/HCI project flagging attention and interaction patterns in a session
- Peet-to-Peet — Uber-style UI concept for home-cooked food
27+, spanning cloud (HCCDA AI + Cloud), SQL/data, ML fundamentals, and a Micro MBA — full list on LinkedIn.