A note about your information

Privacy, in plain words.

Joining the Moa waitlist gives us a way to tell you when early access is ready. Here’s what happens to the information you share.

What we collect

We store your email address, selected language, a broad referral category (such as Reddit, without the individual community), and the time you consented and joined. You don’t need to share your name, files, or payment details. For abuse prevention, we also use temporary pseudonymous identifiers derived from an IP address and time window. The original IP is not stored in the waitlist store; these identifiers expire after 10 minutes and are cleared by daily cleanup.

Why we use it

We use your registration to send Moa launch announcements and early-access invitations. The referral source helps us understand where interest comes from. We don’t sell your email or use it to send unrelated promotions.

Where it goes

This website is hosted on Vercel. Waitlist records are stored in a private Vercel Blob store and are accessible to the Moa operator and the service providers needed to run the site. Vercel also processes technical request information, such as IP addresses and request logs, to host and protect the service. This is separate from the Mac app’s local file processing.

How long we keep it

Waitlist records expire 365 days after signup and are deleted in the next daily cleanup. If you cancel earlier, your record is deleted at cancellation. Technical hosting logs follow the hosting provider’s retention settings.

How to leave

After joining, you receive a private cancellation link on the confirmation screen. No confirmation email is sent automatically, so save the link somewhere safe. Opening that link lets you confirm removal of your waitlist registration. Anyone with the link can cancel that registration, so don’t share it publicly.

What stays in your browser

We remember your language and analytics consent choice in local browser storage. Only after analytics consent, we also store a random analytics identifier that is valid for 90 days and is not reused after expiry. We do not store your email in browser storage. Your private cancellation link carries a token in the URL fragment so it isn’t included in ordinary page requests.

Optional usage analytics

Usage analytics are off by default. Only when the analytics service is configured and you explicitly opt in do we send fixed events to Amplitude: landing visits, language changes, signup attempts and outcomes, and download clicks after release review. Events include language, a broad referral category, session start time, and a random identifier. Failure reasons use fixed categories. We do not distinguish individual Reddit communities.

We never send email addresses, file or folder names, paths, full URLs, query strings, fragments, cancellation tokens, free text, raw errors, or full referrers. We do not use the Amplitude SDK, automatic click capture, or session replay. Your browser sends only to this site's collector, which does not forward your original IP address to Amplitude. For abuse prevention, the collector temporarily uses a pseudonymous identifier derived from your IP and a one-minute time bucket, held only in server memory. Vercel's technical request logging still applies.

Use the analytics settings below to allow, decline, or withdraw. Declining or withdrawing immediately stops collection on this page and removes the saved permission and analytics identifier when browser storage permits. If the browser blocks storage changes, a previous permission may still apply on a later visit; the settings display the storage failure and how to clear this site’s saved data. It does not undo events already sent. Global Privacy Control or Do Not Track disables collection regardless of your saved choice. Cancellation pages do not collect usage analytics.

The random identifier is created only after consent. It is valid for 90 days and is not reused after expiry; expiry does not physically erase browser storage while the site is closed. If storage is unavailable, it exists only in the current page's memory. A new page or 30 minutes of inactivity starts a new session. Data already sent to Amplitude follows the configured project's retention policy. If analytics are not configured, no events or identifiers are created and the settings show that state.

And the Mac app?

Basic sorting and text recognition run locally on your Mac. Optional AI review sends the file and folder names you approve through your installed Codex or Claude Code CLI. It doesn’t send document contents or images. The AI provider’s own terms and account limits apply.

The Mac app has its own optional analytics consent in Settings, separate from this website. If you opt in, it sends fixed action names, app version, language, selected operation categories, result and count/time ranges, session start time, and a random identifier to Amplitude through our collector. It never includes file names, paths, contents, email addresses, category names, or raw errors. The app identifier is valid for 90 days and is not reused after expiry; an inactive app checks expiry when it next runs. Withdrawing in the app stops future collection and removes that Mac’s identifier, but does not delete events already sent. Website and app identifiers are separate. Connection providers may process technical logs.

Feedback from the Mac app

Sending feedback is optional and separate from analytics. Only after you choose Send do we receive the type of feedback, the message you wrote, your optional reply email, app version, language, a random submission reference, and the receipt time. The app does not attach files, filenames, paths, document contents, diagnostic logs, or an analytics identifier. Please leave confidential information out of your message. We use feedback to improve Moa and, if you provide an email, to reply about your feedback; it does not subscribe you to marketing.

Feedback is stored in private Vercel Blob storage for the Moa operator to review. It expires after 90 days and is removed in the next daily cleanup. For abuse prevention, short-lived pseudonymous identifiers derived from an IP address and time window are stored separately; raw IP addresses are not included in feedback records. Vercel may process technical connection logs under its own retention settings. A successful submission shows a reference number; email receipts and replies are not sent automatically.

Last updated: November 6, 2026.