Building a Knowledge Base from Scratch

Upload your documents

Start by gathering the files you already have — PDFs, Word docs, spreadsheets, notes from other apps. Upload them to the Documents section. Chunk indexes each file so the AI can search through them semantically.

Tip: Tag documents as you upload them. A simple taxonomy like "Work", "Personal", "Research" helps later when you want to filter.

Create notes from conversations

As you chat with Chunk — whether asking questions about your documents or exploring new topics — save useful AI responses as Notes. Each saved response becomes a clean markdown note you can edit and organize.

Tip: Use note templates to structure your notes consistently. Research Summary works great for captured insights.

Connect notes with wiki-links

As your note collection grows, start connecting related ideas. Type [[ in any note to link to another note. This creates a web of knowledge where you can navigate between related concepts. Backlinks automatically show you which notes reference the one you're reading.

Explore the Knowledge Graph

Switch to Graph View in Notes to see all your notes and their connections as a visual map. Look for clusters of densely connected notes (strong topic areas) and isolated notes (potential gaps). This bird's-eye view helps you identify what to explore next.

Organize with Collections

Group related documents, notes, and web content into Collections organized by project or topic. Each collection becomes a focused workspace you can chat with, getting AI responses that draw from exactly the sources you've curated.

Tip: Save important notes to Documents (via the hard drive icon) so they're also searchable in collection-scoped chat.