Greetings,
We are currently exploring the implementation of GraphRAG using HANA Cloud / Datasphere Knowledge Graph. We are looking to leverage SAP AI Core (Generative AI Hub) for this purpose and are in the process of gathering information on the specific implementation methods. However, we are currently facing a lack of detailed information and would greatly appreciate your expertise and insights.
Specifically, we would be grateful if you could provide guidance on the following points:
Specific Implementation Methods for GraphRAG using SAP AI Core (Generative AI Hub):
To realize GraphRAG, which functionalities of SAP AI Core should we utilize, and how can we effectively integrate them with HANA Cloud (Datasphere) Knowledge Graph?
Could you provide detailed information such as concrete architecture examples, implementation steps, and considerations?
For instance, what components (models, pipelines, etc.)
should be built within AI Core, and how can we optimally orchestrate query execution against the Knowledge Graph, vector search, and response generation?
Comparison of Architecture with Microsoft's GraphRAG:
As a general concept for GraphRAG, there is an architecture proposed by Microsoft, as referenced here: https://microsoft.github.io/graphrag/
(Considering aspects such as Community Report generation, Global Search, and Local Search).
Is the GraphRAG architecture within the SAP environment similar to Microsoft's architecture?
If there are differences, could you please elaborate on SAP's unique architecture or the optimized GraphRAG approach for the SAP environment?
In particular, we are interested in understanding the unique advantages and characteristics of SAP's GraphRAG when leveraging Knowledge Graphs.
We look forward to receiving your valuable information and opinions.
Thank you in advance for your assistance.
Bluesky
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.