Skip to main content
search

Graph Power Hour Ep1

GraphRAG:

The Good, the Bad, and the Ugly

Webinar Series with Paco Nathan & Featuring Dr. Clair Sullivan

Watch the second podcast of Paco Nathan’s Graph Power Hour. This week’s topic – GraphRAG: The Good, the Bad & the Ugly. 

ABOUT THIS EPISODE

GraphRAG is having a moment! Thanks to the GenAI revolution, more people than ever are turning to graphs to solve problems like retrieval-augmented generation (RAG) and minimize hallucinations common in GenAI models. GraphRAG structures excel at representing complex, interconnected data—critical for improving the contextual understanding that generative AI requires. By leveraging graphs, AI systems can better discover relationships, link entities, and enrich data, making them more effective at handling real-world complexity. As AI evolves, graphs are emerging as key enablers for more sophisticated, context-aware insights.

However, applying graphs to back GenAI applications is not without challenges. Extracting nodes and relationships from unstructured text via natural language processing (NLP) is difficult, as language is rich in context, idiomatic expressions, technical jargon, and implicit meanings that don’t neatly fit into graph structures. While large language models (LLMs) reflect general language usage, they often fail to capture subtle nuances and technical details, leading to oversimplified or inaccurate graph representations. This disconnect between the richness of language and the structured nature of graphs can hinder a system’s ability to generate significant insights.

In this talk, we will explore the use of graphs “from the trenches,” examining real-world challenges encountered when applying them in complex systems. While graphs offer immense theoretical potential, translating that into practical applications often reveals unforeseen difficulties. From handling incomplete, ambiguous, or unresolved data to scaling large, interconnected datasets, the journey is rarely straightforward. We’ll also discuss the challenges of integrating graphs with AI models, where the structured nature of graphs can clash with the messiness of real-world information. These practical insights will highlight the gap between the idealized use of graphs and the realities faced in the field.

Resources

Get the slides here

Visit Dr. Clair Sullivan’s Github here

Read the Neo4 blog featuring: SafeGraph + WHISARD + PPP here

Read the guest blog for Senzing here

Dr. Clair Sullivan’s tutorial here

NEo4j Graph Builder

Retrieval-Augmented Generation for AI-Generated Content: A Survey

A Benchmark To Understand The Role Of Knowledge Graphs On Large Language Model’s Accuracy For Question Answering On Enterprise SQL Databases

We’d love to see you at our next event! Keep an eye out for our upcoming webinars by subscribing to our mailing list.

Paco nathan data scientist senzing

Paco Nathan
Principal DevRel Engineer

Paco Nathan leads DevRel for the Entity Resolved Knowledge Graph practice area at Senzing and is a computer scientist with +40 years of tech industry experience and core expertise in data science, natural language, graph technologies, and cloud computing. He’s the author of numerous books, videos, and tutorials about these topics.

Close Menu