Memgraph, the open-source in-memory graph database platform, is breaking new ground with the release of two major tools designed to simplify and democratize access to Retrieval-Augmented Generation (GraphRAG) technology. The AI Graph Toolkit, available immediately, and an upcoming MCP Client aim to make it easier for developers to bridge traditional SQL and unstructured data with the advanced reasoning capabilities of graph-based AI systems.

Opening the Door to GraphRAG for All

GraphRAG—first introduced by Microsoft in 2024—has quickly become one of the most effective methods for improving large language model (LLM) reasoning. By combining graph databases with retrieval-augmented generation, GraphRAG helps LLMs deliver accurate, context-rich, and non-hallucinatory answers.

Despite its promise, GraphRAG adoption has largely been confined to graph database specialists. The vast majority of developers still work within SQL environments, and much of enterprise data remains trapped in relational and unstructured formats. According to Gartner, 62% of global DBMS revenue in 2023 came from relational systems, while 80–90% of new enterprise data is unstructured.

Memgraph’s new toolkit and client aim to close that gap.

Automating Data Transformation with the AI Graph Toolkit

The AI Graph Toolkit is a set of open-source libraries and utilities that automate the conversion of SQL and unstructured data into a graph format ready for GraphRAG pipelines. The toolkit includes SQL2Graph and Unstructured2Graph, both of which allow developers to seamlessly build knowledge graphs from existing data sources.

This automation replaces the tedious manual work traditionally required—mapping tables, identifying entities, performing entity resolution, and converting unstructured files into vector embeddings. Developers can now prepare data for GraphRAG-driven applications with 10x faster development cycles, based on results from multiple beta test sites.

Marko Budiselić, Memgraph’s CTO and Co-founder, emphasized the breakthrough:

“Our new tools make GraphRAG development far more accessible. Developers no longer need deep graph expertise to transform SQL and unstructured data into powerful AI-ready knowledge graphs. The result is faster, smarter, and more reliable AI chatbot performance.”

The MCP Client: Expanding Graph Connectivity

Launching later this month, Memgraph’s MCP Client will further streamline integration between graph systems and diverse data environments. Fully compatible with the emerging standard for context engineering, the client enables seamless connectivity between Memgraph’s graph capabilities and various back-end sources through other MCP Servers.

This interoperability is expected to accelerate AI developer productivity, especially for organizations adopting hybrid data architectures.

From Prototype to Production with JumpStart

To help enterprises deploy production-ready GraphRAG solutions quickly, Memgraph has introduced the JumpStart Programme. The initiative combines a Memgraph Enterprise license with over 20 hours of expert support and implementation. Within weeks, companies can achieve a full GraphRAG pipeline featuring hybrid search (vector + graph reasoning), sub-second latency, and explainable, provenance-rich answers.

The AI Graph Toolkit is now available for download via GitHub and is also accessible through Memgraph Cloud. More details about the Toolkit and MCP Client will be shared at Connected Data London, held at the Leonardo Royal London Hotel Tower Bridge on November 20–21, 2025.

Memgraph’s push into GraphRAG accessibility signals a significant leap for the AI and data management landscape—one that could finally unite structured and unstructured data under a single, intelligent framework for real-time enterprise applications.

Get in touch with our MarTech Experts.