This session explores how graph-based methods and large language models can complement one another - combining structured relational reasoning with the flexibility of language understanding. Discover emerging architectures and use cases where the fusion of these approaches unlocks new levels of performance and interpretability. Key learnings: → Understand how graph structures can enrich LLM inputs, outputs, and training signals → Explore hybrid model designs that leverage both symbolic and neural reasoning → Learn from real-world applications where graph-LLM integration drives measurable gains