Learn to develop and train custom Graph Neural Networks for binary function search in this 38-minute conference talk from 44CON Information Security Conference. Discover how to leverage machine learning techniques to analyze binary code structures, understand the fundamentals of graph-based representations for binary functions, and implement neural network architectures specifically designed for reverse engineering tasks. Explore practical approaches to training models on your own binary datasets, examine the challenges of feature extraction from compiled code, and gain insights into improving binary analysis workflows through automated function identification and similarity detection using deep learning methods.