Privacy

Without an account: your chosen energy level and reading-comfort preference live in your own browser’s local storage. They never leave your device, and clearing your browser data removes them completely.

With an account: we store what you give us — your email and sign-in details (handled by our authentication provider, Clerk), your conversations, memory cards, plans, notes, energy check-ins, your rests and your own Recovery activities, the name and preferences you set, and the energy level you last chose with the time you chose it. If you subscribe, we also keep the customer number and subscription status our payment provider reports. If you switch on the morning email, your email address is used to send it, once a day, through Cloudflare, and every one carries a link that stops it. This is the data Nami needs to be useful to you across devices and across days. If you connect a calendar, we keep what opens it (the link, the app password or the Google or Microsoft sign-in, locked with a key the database does not hold) and its entries for the next two months; today’s and tomorrow’s entries are part of what the AI model receives when you chat. Pictures you add for Today are kept for you alone. Pictures you send in chat are stored with Cloudflare R2, where only your account can open them; they are deleted when an edit takes them out of a conversation, and when you delete your account.

We never sell your data — not today, not in a pivot, not in the fine print. We never run hidden energy scoring: there is no model watching your behavior to guess how you feel. We never use your memories in chat unless you have explicitly enabled that. We never diagnose, infer a medical condition, or present Nami as emergency or medical care.

Energy check-ins, stress readings, your rests and your own Recovery activities are personal in a way most app data is not. We treat them accordingly: stored only to serve you, visible only to you, never used for advertising, and never shared with third parties beyond the processors required to run the service (hosting, authentication, payments, email delivery and storage for the pictures you send in chat, both through Cloudflare, and the AI model that answers your chats).

You can add Nami to Claude, Anthropic’s assistant, as a connector; Settings gives the address. You sign in to Nami to allow it, and from then on Claude can read your board with each task’s steps, the plans you kept, your notes, Recovery, the calendars you connected, the level you set and what Today suggests, and your memories only while you let them surface in chat. When you ask it to, Claude can also add and change things in Nami: tasks and their steps, notes, memories, your rests and your own Recovery activities, your level, and entries in the calendar you chose for Nami to write to; and it can delete a note, a memory or one of your own Recovery activities. Claude asks you before it changes anything. Nami’s “Ask before Nami writes” setting covers what Nami does, not Claude. What Claude reads becomes part of your conversation there, which Anthropic handles under its own terms. Removing the connector in Claude ends it.

Under the GDPR and similar laws, you can access the data we hold about you, correct it, export it, or ask us to delete it. Settings › Your data downloads all of it as one file, whether or not you subscribe. For anything else, email privacy@kumbat.io and a human will handle it — no dark patterns, no retention games. The controller of your data is Entro314 Labs Single Member P.C., in Greece. Deleting your account deletes your data from our active systems.

We disclose personal data to a government, a court or the police only when the law requires it. Every request is checked for whether it is lawful; one that is not, or that asks for more than it needs, we challenge. When we must answer, we hand over the least that answers it. We tell you about the request unless the law forbids it, and we keep a record of every one.

Authentication is provided by Clerk. AI responses are generated by third-party language models, which receive the conversation context needed to answer you. Nami runs on the open Kumbatio energy system — the SDK is inspectable, and the pledge behind it lives at kumbat.io. Errors are written to our own server log, without what you wrote, your memories or Nami’s replies. Payments, if you subscribe, are handled by Polar; we never see your card.

Dictation in chat listens only while the microphone button is on. The recording goes to our server and on, through the Vercel AI Gateway, to an OpenAI transcription model that turns it into words; the words come back into the message box, and nothing goes to Nami until you send it. We keep neither the recording nor the words.

Reading a reply aloud uses the voices already on your device, through your browser; the words are not sent to us or anyone else for it.

If this policy changes in a way that matters, we will say so plainly in the app before it takes effect. The date at the top always reflects the current version.