Central Knowledge Base

Feed global requirements templates and historical reference guides to improve Context Retrieval and Analysis.

Index New Reference Document

PDF, DOCX, TXT, Excel (up to 10MB)

Example Files

Download sample specifications and templates to test the system indexing and retrieval features.

Indexed Reference Documents

0 items

Loading knowledge base...

How vector indexing works: Documents are automatically parsed (extracting plain text from formats like PDF/Word/Excel), split into ~800 character chunks with overlap, embedded using Google Gemini embeddings, and stored in a vector database index. During analysis, the Context Retrieval Agent queries this index to find the most relevant templates and rules.