My notes were dying in folders, so I built an infinite canvas that keeps them arguing

Description and use

Knovum replaces folder storage with an infinite canvas that keeps sources and their relationships, tracks citation validity, and runs AI tools locally, improving knowledge workflow.

Problem

Desktop folders fail to capture the evolving, interrelated state of knowledge, causing loss of context and stale references.

Outcome

Knovum replaces folder-based storage with an interactive canvas that preserves source relationships, tracks citation validity, and integrates local AI tools, reducing stale references and improving knowledge workflow.

Workflow

• 1. Import documents, clips, and media as cards onto the canvas.
• 2. Group, tag, and create named views for different lenses.
• 3. Use typed knowledge objects (Concept, Question, Claim, Evidence, Experiment, Decision) and relation kinds to model relationships.
• 4. Run AI tasks (transcription, summarization, etc.) locally via the Task Center.
• 5. Generate derivations that record inputs, revisions, and model used.
• 6. Monitor Knowledge Health for stale evidence or unsupported claims.
• 7. Export or share a .knovumspace file containing all content and relations.

Evidence

I counted my desktop folders last month. Forty-three. The naming scheme had evolved from "research" to "research-sorted" to "research-sorted-v2" to "final-actual".

Source

Folders have no slot for any of that. Folders have "inside". They don't have "between".

Source

I built Knovum: a local-first desktop knowledge workspace where your sources sit on one infinite canvas, and the relationships between them are first-class citizens.

Source

Knovum ships six typed knowledge objects — Concept, Question, Claim, Evidence, Experiment, Decision — connected by ten relation kinds: supports, contradicts, depends-on, supersedes, answers, derives-from, and friends.

Source

Knowledge Health continuously reports claims with no evidence, accepted-but-contradicted claims, evidence whose source document changed under it, open questions that already have answers.

Source

Ask answers from the sources you select, and every citation keeps its location: a page in a PDF, a time range in a recording, a heading and character span in a note, a line range in code.

Source

All running locally after a one-time model download, SHA-256-pinned weights, Metal-accelerated on Apple Silicon.

Source

Select material, run an intent — generate document, image, or video; remove background; split layers; add subtitles. Every result records its derivation: the inputs and their revisions, the action that ran, the engine or model that executed it.

Source