Zettory
What it gives you
- Hybrid lexical and semantic retrieval with fused ranking
- Search Markdown notes, Kanban cards, drawings, links, visible canvas text, and OCR
- Filter results by source type and scope
- Local embedding index for semantic matching when wording differs
- Open the exact supporting source directly from search results
Search by words and by meaning
Zettory combines normal text matching with local embedding-based retrieval. Exact terms remain useful for names and phrases, while semantic matching can surface related notes even when your query uses different wording.
A knowledge layer over the notes you already have
The index is built from note and workspace artifacts instead of forcing you to rewrite everything into a separate knowledge-base format. Markdown, drawings, link context, board cards, visible canvas text, and OCR can all contribute searchable evidence.
Search is not a second copy of your notes
Results point back to the source artifact. The index exists to improve retrieval while the note or workspace remains the durable place you inspect and edit.