Knowledge Bases
Upload documents, sync cloud folders, and give your agents knowledge.
The Knowledge Bases page is where you manage all the documents, databases, and code repositories your agents can search and reference.
Page layout
Your document library with sub-tabs:
Library — browse all uploaded files
Processing — monitor ingestion pipeline status
Multimodal — images, PDFs, and rich media documents
Search — test semantic search across your knowledge base
RAG — retrieval-augmented generation configuration
Upload — bulk upload interface
SQL and semantic database tools:
SQL Explorer — query your data directly
Semantic Layer — natural language to SQL mappings
Query Templates — saved queries
Training — teach the NL2SQL system
Schema Browser — explore table structures
Audit History — query execution logs
Document templates for standardised content creation. Agents can use templates to generate consistent reports, summaries, and documentation.
Code repository indexing. Connect repos and browse your codebase as a knowledge graph — agents can search code semantically.
Stats bar
Total Documents
Number of files in your knowledge base
Processed
Files that have been chunked and embedded
Storage Used
Total disk space consumed
Vector Chunks
Number of searchable embedding chunks
Quick start
Click Upload Documents in the top right
Drag and drop files (PDF, MD, DOCX, TXT, and more)
Wait for processing (chunking + embedding)
Test with a search query in the Search sub-tab
Your agents can now reference these documents in chat
Documents are automatically chunked using semantic boundaries and embedded for vector search. No manual configuration needed.
Subpages
Uploading, browsing, and managing documents
Syncing folders from Dropbox and cloud storage
SQL Explorer and semantic querying
Document templates
Code repository indexing
Last updated