A powerful AI agent needs access to the right data. In this video, you’ll learn how to use knowledge bases in StackAI to give your agent access to private files, systems, and APIs—like SharePoint, Google Drive, Notion, and more. You’ll see how to configure file indexing, set search filters and metadata, control access with roles-based permissions, and let your agent search data only when needed using tools.

Summary

• Why knowledge bases matter: build context-aware agents powered by private data

• Two ways to configure knowledge bases: left-side panel and within workflows

• Upload static documents (e.g. PDFs) and adjust chunking, indexing, and search settings

• Connect knowledge bases to LLMs, referencing documents and questions inside prompts

• Preview parsed files and assign metadata tags for better document discovery

• Drag and drop other types of nodes: APIs, tables, web sources, or dynamic file systems

• Connect to SharePoint, Google Drive, Dropbox, and more via secure credentials

• Enable auto-sync to keep data up to date as files change

• Use a knowledge base as a tool, so the LLM queries it only when needed

• Configure search strategy, result length, and advanced tool settings

• Set up roles-based access control to restrict data access by person or group

• Share knowledge bases securely via the central manager interface

• Next up: deep dive into LLM tooling and advanced capabilities inside your workflows