What AI tools can analyze personal notes for patterns?

The landscape of note analysis tools

General-purpose LLMs — ChatGPT, Claude — can analyze pasted text and surface observations about your thinking. But they produce one-off outputs with no structure, no persistence, and no way to compare your patterns against other thinkers. Each conversation starts fresh.

Obsidian plugins like Smart Connections do local semantic search within your vault — useful for navigating existing notes, but they don't extract cross-document patterns or produce a structured picture of how you think. They surface proximity, not structure.

Rodin is purpose-built for cross-document pattern extraction. It reads your entire body of writing at once, identifies what recurs across years rather than sessions, and outputs a structured, shareable intellectual fingerprint. The result can be matched against other thinkers in the network.

What Rodin extracts — the 8 layers

1. 01

   One-liner

   A single sentence distilling your entire intellectual project — not what you do, but what you are trying to figure out.

2. 02

   Recurring themes

   The topics and tensions that reappear across your writing regardless of what you think you are writing about.

3. 03

   Open questions

   Questions your notes circle without resolving — genuine unresolved tensions you keep returning to across years.

4. 04

   Mental models

   The conceptual frameworks you reach for when explaining or analyzing something. Every thinker has a private toolkit; the fingerprint surfaces yours.

5. 05

   Intellectual DNA

   The thinkers, texts, and traditions that shaped how you reason — whether or not you cite them directly.

6. 06

   Blind spots

   Patterns of avoidance: questions you do not ask, perspectives you do not consider, tensions you resolve too quickly.

7. 07

   Core question

   The single question underneath all your other questions. Every thinker, examined closely enough, is trying to answer one thing.

8. 08

   Archetype

   A three-word identity capturing how you relate to ideas — something like "The Contrarian Synthesiser" or "The Patient Systematizer."

Beyond pattern extraction: Cognitive Topology Analysis

Alongside the 8-layer fingerprint, Rodin runs a 12-dimensional Cognitive Topology Analysis (CTA) on your writing. This is a proprietary engine built from pure linguistic computation — no external API calls. It captures not what you think, but how you structure thought.

The 12 dimensions include epistemic confidence (how certain your writing sounds), temporal orientation (whether you reason from past patterns or future possibilities), argument density (how many distinct claims you make per unit of text), and conceptual leap distance (how far you jump between ideas without explicit bridging).

Other dimensions measure whether you reason from authority or first principles, whether you think in abstractions or concrete examples, your dialectical complexity, and your intellectual tempo. Together they produce a 12-dimensional cognitive signature — enabling structural comparison between any two thinkers in the network.

What makes note analysis different from chatting with AI

When you chat with an AI about a topic, you get that AI's response to what you said in that session. When you submit your full vault, the AI sees your entire intellectual history at once. It sees what you return to across years, not what you think about on one day.

Cross-document visibility changes what becomes visible. A single note might mention a theme once. Thirty notes mentioning it unprompted, across different months and contexts, signals something structural about how you think — not just what you happened to write about.

Rodin reads for recurrence, not for content. The intellectual fingerprint it produces describes the architecture underneath your writing — the questions and frameworks that shape your thinking whether or not you are conscious of them.

Supported note-taking tools

Rodin accepts any text or markdown. For Obsidian, you can upload an entire vault folder of .md files at once — Rodin parses wikilinks and extracts knowledge graph structure as additional context. For Notion, export your workspace as markdown and paste or upload the result.

Logseq, Roam Research, Bear, Apple Notes, and any other tool that can export plain text or markdown all work the same way. If you can copy your writing into a text field, Rodin can analyze it.

Frequently asked questions

How much text do I need?

A few thousand words is a workable minimum. More text — and more variety across time — produces a more accurate fingerprint. Rodin is looking for what you keep returning to across your whole body of writing, not individual pieces. An Obsidian vault of a year or more tends to yield the sharpest results.

Is my writing used for AI training?

No. Your writing is not used to train AI models. A portion is stored to support your profile — it is never used to train any AI model. You can delete your profile and all associated data at any time using your management link.

Can I see what the analysis looks like?

Yes. Browse example profiles on the Discover page to see full intellectual fingerprints — the 8-layer analysis, cognitive topology visualization, and match scores. Each profile page shows what the output looks like before you commit your own writing.