Mjara Docs
Tools Reference

Knowledge & RAG Tools

Tools for knowledge base management and semantic search.

ToolDescription
knowledge_base_searchSemantic search across indexed documents
document_indexerIndex documents for RAG retrieval
policy_assistantQuery company policies using natural language

Search your company knowledge base using natural language:

  • Semantic search powered by vector embeddings
  • Returns the most relevant documents ranked by relevance
  • Configurable number of results (default: top 5)

Document Indexer

Index documents for the knowledge base:

  • Upload and index PDFs, Word documents, and web URLs
  • Automatically chunk and embed document content
  • Update indexed documents when source files change
  • Remove documents from the index

Policy Assistant

Answer questions about company policies:

  • Query HR policies, operational procedures, and guidelines
  • Provide accurate answers with source references
  • Handle follow-up questions with conversation context

On this page