Mohamed Magdy

I’m a Data Scientist focused on turning AI into reliable, production-ready products. I architected and deployed a real-time conversational AI web application using OpenAI GPT-4o-mini and WebRTC, supporting 500+ concurrent users with sub-2 second response times.

In that project, I owned the end-to-end lifecycle—from requirements gathering through production deployment under tight deadlines. The resulting browser-compatible experience increased client engagement by 150%.

I also completed the IBM Data Scientist track covering Python, SQL, and R, applying ML tools to large datasets. My builds include a Flutter stock predictor/trading simulator (65% accuracy with LSTM and Gradient Boosting) and an Arabian music lyrics generator trained on 1,000,000+ lyrics, achieving an 85% coherence score.

For applied AI engineering, I developed a medical AI chatbot using RAG with LangChain orchestration and Pinecone vector retrieval, served via a Flask REST API for source-backed clinical answers. I’m comfortable across the stack, using Git/GitHub, REST APIs, and frameworks like Flask and Django, while continuing to deepen my ML and data science capabilities.