About
Notes shouldn't feel like homework.
Kanto Notes was born out of one recurring frustration: the best ideas hit at bad moments, and the moment you try to file them properly you lose the thought.
Every existing notes app forces you to be the librarian: pick the folder, invent the title, tag it, remember where you put it, and hope you can find it again by exact keyword months later. That work happens at the moment of capture — exactly when the idea is most fragile.
We built Kanto on a single bet: LLMs are cheap and good enough now to do the librarian's job in the background. The user should only ever have to do two things — dump the thought (voice, text, whatever's fastest) and ask for it back (in plain language, whenever it's needed). Everything between those two moments — titling, categorizing, tagging, indexing, retrieving — is machine work.
Principles we build by
- 01
Capture is sacred.
The home screen is a textarea and a mic button. It never grows. Every new feature must earn its way onto a different screen.
- 02
AI is invisible until it isn't.
Users never wait for AI on the capture path. Enrichment runs in the background; the note is saved instantly.
- 03
Retrieval beats organization.
Folders and manual tags are opt-in. If the AI does its job, users never need to file anything by hand.
- 04
The user's data is their data.
Bring-your-own-key from day one. Export is a first-class feature, not a churn-prevention tactic.
- 05
Explain the AI's work.
Auto-tags, categories, and answers always show what the model did and let the user override.
Who it's for
The person who thinks faster than they type, has thoughts in inconvenient places (walks, trains, showers, meetings), and has tried three note apps and abandoned all of them because filing was harder than remembering.
If your current setup is Apple Notes plus a Google Doc named dump.md, Kanto is for you.