Operationalizing GraphRAG: Lettria’s Scalable Architecture with Neo4j and Qdrant

Learn how Lettria designed a production-scale GraphRAG system by integrating Qdrant’s high-performance vector engine with Neo4j’s graph-based knowledge representation. 

Qdrant needs the contact information you provide to us to contact you about our products and services. You may unsubscribe from these communications at any time. For information on how to unsubscribe, as well as our privacy practices and commitment to protecting your privacy, please review our Privacy Policy.

 

GraphRAG is expanding what's possible with RAG, enabling systems to combine the ability to capture the semantics of vector search with the structure of knowledge graphs for greater context and control and better understanding of the relationships in your data.

In this session, you’ll learn how Lettria designed a production-scale GraphRAG system by integrating Qdrant’s high-performance vector engine with Neo4j’s graph-based knowledge representation.

The result: a platform capable of handling over 100 million embeddings with sub-200ms latency and delivering 20–25% higher accuracy than traditional RAG in real-world applications.

We'll talk about Qdrant's universal query API, more about Lettria's architecture and approach to GraphRAG.

layout

Speakers:

1
Romain Albrand, AI Engineer at Lettria
Lettria is a document intelligence platform specifically designed for regulated industries, offering a robust solution for organizations that require meticulous handling of sensitive information. This platform leverages advanced technology to enhance the accuracy and reliability of document processing, making it an essential tool for enterprises that manage high-stakes content. Lettria’s unique GraphRAG technology significantly boosts accuracy by 30% compared to traditional RAG methods, ensuring that users can trust the outputs generated from their documents.
2
Kacper Lukawski, Snr Developer Advocate at Qdrant
Qdrant is the open-source vector search engine that powers the next generation of AI applications. Built for performance and scalability, Qdrant enables real-time, high-dimensional similarity search at scale, making it ideal for production-ready semantic search, recommendation systems, and RAG.
3
Jennifer Reif, Developer Advocate at Neo4j
Neo4j is the world’s leading graph database and analytics platform, helping organizations uncover relationships and patterns in connected data. It is trusted by Fortune 500 companies and developers alike to power intelligent, real-time applications.