About the Customer 360 Copilot Demo

The goal of this site is to demonstrate how CoPilot works on a sample web storefront dataset. There are many benefits to combining large-language models (LLMs) work with graph databases.

Here are just some of the key benefits of combining LLMs and graph databases.

Mapping Questions to Graph Queries (TigerGraph InquaryAI)

LLMs can translate complex natural language questions into graph database queries. This helps users access information intuitively without requiring in-depth technical knowledge, allowing them to interact with graph data structures efficiently.

Mapping Questions to Document Sections (TigerGraph SupportAI)

Generative AI can identify relevant sections within large documents that relate to a user's question, streamlining the information retrieval process. This capability enhances the ability to navigate extensive datasets and extract specific insights without manual searching.

Enhanced Knowledge Representation through Concepts in Taxonomies and Ontologies

Integrating taxonomies and ontologies with graph databases allows LLMs to understand domain-specific relationships, improving the accuracy of results. This structured knowledge aids LLMs in comprehending context and ensures that the generated content remains truthful and logically coherent.

Improved Data Connectivity and Context

Graph databases excel at representing interconnected data. By mapping LLM outputs to graph structures, relationships between entities become clearer, leading to a more holistic view of the information. This contextual understanding enhances the relevance and coherence of generative outputs.

Reduced Ambiguity and Increased Accuracy

Taxonomies and ontologies provide a structured framework that reduces ambiguity in responses. This structure allows generative AI to focus on specific entities and relationships, resulting in answers that are more precise and aligned with domain-specific knowledge.

Customizable and Domain-Specific Applications

Combining LLMs with graph databases enables the creation of tailored applications, where specific ontologies define the context and relationships. This customization allows for domain-specific responses and insights, catering to particular industry or research needs.