Ahmed Elzayat

Developed an intelligent voice assistant leveraging a Retrieval-Augmented Generation (RAG) pipeline for real-time, context-aware conversational AI, integrating dynamic location, weather, date/time and web search services.

Built an in-house medical imaging platform, featuring advanced AI segmentation and LLM-based reporting capabilities through a model designed for generalizability across various brain tumor populations.

Developed a multilingual NER NLP system using mBERT and XLM-RoBERTa transformers, evaluated on the PAN-X dataset. Compared monolingual and multilingual fine-tuning strategies across German, French, Italian, and English.