RAG System Administration

What problem does it solve?

This skill helps administrators efficiently manage and optimize the Retrieval-Augmented Generation (RAG) system that powers knowledge-base searching, retrieval, and indexing. It addresses the complexity of maintaining high-quality search results across large document collections.

Core Features & Use Cases

• Reindex the vector store to refresh embeddings after model or data changes.
• Optimize search parameters to improve relevance, recall, and performance.
• Create and manage custom indexes for topic-focused retrieval.
• View and update global RAG settings to align with organizational needs.
• Use cases: a data team that needs faster, more relevant search across thousands of articles; an AI assistant that must reconfigure RAG behavior for a new project.

Quick Start

View current RAG settings with view_settings(section="rag"), then adjust parameters with update_settings(section="rag", updates={...}), reindex the collection with reindex_collection(force=true, batch_size=50), and monitor progress with get_task_status(task_type="reindex").