The biggest AI challenge is not model architecture — it is data fragmentation. At United Nations, Batoul processed hundreds of terabytes of fragmented human capital data from regions-in-transition, fusing Geospatial AI, Graph Neural Networks (GNNs), and Large Language Models (LLMs) to produce interactive geo-enabled dashboards that turned fragmented human capital signals into actionable policy intelligence across society, economy, and environment.

Batoul works at intersections others avoid. Her thesis fused Graph Convolutional Networks (GCNs) with multi-atlas ensemble regression on functional MRI (fMRI) data to predict biological brain age beyond what either approach achieves alone. At the UN, GNNs + LLMs + Geospatial analytics forecast labour-market demand in fragile economies. As consultant: same ideas, carried from whiteboard to production MLOps and a live Retrieval-Augmented Generation (RAG) chatbot.

Beyond the technical: Batoul teaches Science, Technology, Engineering, Arts & Mathematics (STEAM) robotics and 3D-printing to children and trains women and university students in Python, Machine Learning (ML), and Prompt Engineering at vocational centres in Lebanon. Every model built, every dashboard shipped, every student taught is the same argument — AI is not an end in itself. It is a lever for empowering communities and advancing the humane-in-humanity.