Highlights
- Automatically load ontologies from knowledge graphs into GraphRAG-SDK.
- Seamlessly connect structured data to LLMs for querying workflows.
- Simplify pipeline creation without manual ontology setup or deep KG expertise.
We’re excited to announce the release of GraphRAG-SDK 0.5, designed to make working with knowledge graphs (KGs) and large language models (LLMs) more seamless and developer-friendly. If you’ve ever struggled with manually defining ontologies or connecting your structured data to an LLM pipeline, this update is for you.
How It Works
Here’s how the new workflow compares to older methods:
Before: You had to manually create a KG, define its ontology, save it separately, and integrate it into your application.
Now: Simply load your KG (or create one using the SDK), and the ontology is automatically loaded from your knowledge graph and connected to your pipeline.
Previously, integrating a KG into an LLM workflow required manual ontology creation, storage, and connection. This process was tedious and error-prone, especially if you lacked deep domain knowledge in KG structures. For developers managing structured data or pre-existing KGs, the overhead of manually defining ontologies slowed down experimentation and deployment.
With GraphRAG-SDK 0.5, we’ve eliminated these bottlenecks. You can now automatically load an ontology from your KG into the SDK, skipping the manual steps entirely. Whether you’re building a KG from scratch or connecting to an existing one, this release simplifies the process so you can focus on querying and extracting insights.
- Automatic Ontology Loading: No need to manually define or save ontologies anymore. If your KG exists, the SDK handles the rest.
- Predefined Knowledge Graph Support: Connect directly to a predefined KG and start querying its ontology immediately.
- LLM Integration: Seamlessly hook up your ontology to an LLM for Q&A workflows—no intermediate steps required.
- Simplified Pipeline Creation: Bring your structured data, generate a KG using the SDK, and start asking questions without needing to understand every detail of the backend.
- Improved Document Processing: A new progress bar tracks document ingestion, giving better visibility into pipeline execution.
Get Started
GraphRAG-SDK 0.5 takes us closer to handling unstructured data better by simplifying structured data workflows first. By automating ontology management and improving usability, we’re making it easier for developers to unlock the full potential of KGs in AI applications.
If you’ve been waiting for a way to make querying your data as simple as chatting with it—this is it.
- Check out the SDK: https://github.com/FalkorDB/GraphRAG-SDK
- Check out the Colab: https://colab.research.google.com/drive/18iobwZwPAgdBFm3rMT73xs13u4VarN6S?usp=sharing
What is new in GraphRAG-SDK 0.5?
How does it simplify workflows?
Who should use it?
Build fast and accurate GenAI apps with GraphRAG SDK at scale
FalkorDB offers an accurate, multi-tenant RAG solution based on our low-latency, scalable graph database technology. It’s ideal for highly technical teams that handle complex, interconnected data in real-time, resulting in fewer hallucinations and more accurate responses from LLMs.