Mohamed Saber

I've evaluated LLM-generated outputs across CityGPT and an AI-powered social media management agent, assessing accuracy, clarity, reliability, and alignment with task objectives.

For CityGPT, I helped build a chat-based smart-mobility platform that turns natural-language requests into traffic simulation scenarios. I integrated Gemini AI and LangGraph workflows, refined prompts, built FastAPI backend services, and supported PostGIS, OpenStreetMap, and A/B Street integration.

I've also trained and compared YOLO hand-gesture models using annotated Roboflow datasets, evaluating precision, recall, and inference performance. My machine-learning project work includes brain tumor classification with VGG19, movie sentiment analysis, image processing, and recommendation chatbots.

I'm a bilingual Arabic and English Computer Science graduate who brings structured annotation, careful quality review, prompt iteration, and practical Python development to AI evaluation work.