Knowledge Base Integration

GRAI allows you to integrate custom knowledge bases into your geospatial analyses, enabling the AI to reference your specific documents, datasets, and domain expertise.

What is a Knowledge Base?

A knowledge base in GRAI is a collection of documents and information that the AI can access and reference when responding to your queries. This creates a more personalized and domain-specific experience by allowing the AI to:

• Reference your internal documentation
• Use proprietary datasets
• Consider industry-specific terminology and standards
• Apply domain expertise from your uploaded materials

Setting Up Your Knowledge Base

1. Access the Knowledge Base Page

First, click on the Knowledge Base button in the main toolbar:

2. Add Documents to Your Knowledge Base

To add documents to your knowledge base:

1. Click the "Add Document" button
2. Select the PDF files you want to upload
3. Add optional tags or descriptions to help organize your knowledge base

3. View and Manage Your Documents

You can view all the documents in your knowledge base, check their processing status, and organize them by tags:

4. Activate and Use Your Knowledge Base

After your documents are processed and indexed, you can activate your knowledge base for use in analyses. The AI will then be able to reference these documents when responding to your queries.

Supported Document Types

Currently, GRAI supports the following document formats:

Using Knowledge Base in Queries

Once your knowledge base is set up, you can reference it in your queries:

The AI will then combine:

1. General geospatial analysis capabilities
2. Information from your knowledge base
3. Current Earth observation data

Best Practices for Knowledge Base

For optimal results with your knowledge base:

• Use clear, well-structured documents: Well-organized content is easier for the AI to understand
• Include relevant metadata: Dates, locations, and authors help contextualize information
• Break large documents: Split very large documents into smaller, focused sections
• Include visual elements: Charts, maps, and diagrams with clear labels are beneficial
• Update regularly: Keep your knowledge base current with the latest information
• Use descriptive tags: Organize your documents with meaningful tags for better retrieval
• Test with specific queries: Verify the system retrieves the correct information

Limitations

Current knowledge base limitations to be aware of:

• Maximum of 100 documents per knowledge base (enterprise tier allows more)
• Maximum file size of 10MB per document
• Text extraction may not be perfect for complex layouts or scanned documents
• Information is private to your account and not shared with other users
• Processing time depends on document size and complexity

Example Applications

Knowledge base integration is particularly valuable for:

• Environmental consultancies: Incorporating previous assessments and reports
• Urban planning departments: Using city-specific guidelines and historical data
• Research institutions: Leveraging published studies and methodologies
• Disaster response: Incorporating critical local information during emergencies
• Agricultural management: Using crop-specific guidance and historical yields
• Policy enforcement: Applying regulations to geographic contexts
• Historical analysis: Understanding changes over time based on archival documents

For more information on customizing GRAI with your data, see the External Data Integrations guide.