Tuesday, November 12th | From 10:00 AM – 11:00 AM CET
On-line Webinar
Retrieval Augmented Generation (RAG) models are very effective at providing answers quickly by drawing on a large set of documents. However, they struggle to link information from long documents coming from dispersed sources and diverse contexts. This becomes a challenge when responses require a holistic understanding or synthesis of information drawn from multiple documents or passages.
Example: “What is the impact of a supplier change on our product line?”
A standard RAG model would extract relevant passages from documents, but its answer might miss connections between supply chain logistics, inventory data, or product specifications.
GraphRAG, on the other hand, can recognize and map these relationships, giving you a nuanced, interconnected response.
Developed by Microsoft Research, GraphRAG builds upon RAG by integrating knowledge graphs, enabling it to understand data relationships and deliver a more comprehensive answer. It’s meant to enhance specific, complex questions that demand a more global perspective.
What you will learn during the webinar
- Differences between GraphRAG and standard RAG, and what these mean for business insights.
- How the GraphRAG pipeline works to connect and analyze data relationships.
- Sample Comparision of GraphRAG vs RAG answers in real-world examples.
- Benchmarking results that demonstrate performance in various scenarios.
- Potential use cases and considerations for incorporating GraphRAG into your organization.
Registration
Tuesday, November 12th From 10:00 AM – 11:00 AM CET
Online Teams Meeting