Privacy Policy

Four kinds of data:

• Your identity. An account identifier from our authentication provider, your email address, a display name, and profile preferences such as timezone and language.
• Your work. Workspaces you belong to and everything in them: ideas you submit to Validate, questions, analysis reports, saved opportunities, portfolio entries, monitoring rules, notes and research.
• Acceptance records. Which version of these Terms of Use and this Privacy Policy you accepted, and when.
• Operational records. Sign-in sessions, a hashed form of the IP address a session came from, security events such as permission changes, and the request logs our hosting provider keeps, which record the IP address a request came from, the page or endpoint requested and the result.

Separately, the product collects public source material — news articles, public forum posts, package releases, public questions and search-trend data. That material is about the world rather than about you, is shared across all workspaces, and is not treated as your personal data.

To run the service: authenticate you, keep your workspace separate from other people's, produce the analysis you ask for, deliver the alerts you turn on, protect the service from abuse (for example by limiting how many requests can be made in a short time), keep a record of security-relevant events, keep a record of which terms you accepted, and keep encrypted backups so the service can be restored after a failure. We do not sell data, and we do not use it to build advertising profiles.

The legal basis for each purpose is being finalised and will be published here before the service is offered to the general public.

These providers process data on our behalf in order to run the service:

• Authentication provider (Clerk) — holds your sign-in identity and email address.
• Hosting and database (Render) — runs the application, stores its database and keeps its request logs.
• AI model provider (OpenAI) — when you run Validate, receives the text of the idea you submit and the analysis built from it, in order to return the analysis. Discover also sends it text drawn from public sources; that text is not about you.
• Google Trends — when you run Validate, the text of your idea is sent to Google Trends as a search term, to measure interest over time.
• Backups (GitHub Actions and Cloudflare R2) — a copy of the database is made by an automated job on GitHub Actions, encrypted before it leaves that job, and stored with Cloudflare R2. Only we hold the key that decrypts it.
• Email delivery — if, and only if, you turn on email for an alert, the email service configured for this deployment receives your address and the contents of that notification so it can deliver it. No email service is contacted while email delivery is unconfigured, and no marketing email is sent from this product.
• Team chat and webhook delivery — if you turn on a Slack, Discord or webhook channel for an alert, the contents of that notification are sent to the destination configured for this deployment.

The product also reads from public sources such as news feeds, Hacker News, Stack Exchange, GitHub, package registries and search-trend data. Discover sends those sources its own queries, not your data.

The application and its database are hosted in Frankfurt, Germany. Some of the providers above may process data in other countries, including the United States. The safeguards that apply to those transfers have not yet been published here.

This product sets no analytics, advertising or tracking storage. There are no third-party trackers and no marketing scripts, and fonts are served from our own domain. What it does use is limited to what the product cannot work without:

• authentication cookies set by our sign-in provider;
• the workspace you last selected, stored in your browser so the app opens where you left it;
• the identifier of an analysis that is still running, stored for the session so a page refresh does not lose it.

All three are strictly necessary to provide a service you asked for, which is why you are not shown a consent banner for them. If optional storage is ever added, consent will be asked for before it is set.

Validate and Discover analyse ideas and public material automatically, partly with AI models. They assess ideas and markets, not people: the product makes no decision about you that has legal or similarly significant effects.

From Settings → Privacy & data you can, at any time and without asking us:

• Export your data — a machine-readable file containing your account details and the contents of the workspaces you belong to. It deliberately excludes session records, invite tokens and security logs, and it states what it has left out.
• Correct your details — change your display name there; your email address is managed through the account menu by our authentication provider, and the copy we hold is updated from it.
• Delete your account and data — described in full below.

You also have the right to object to or restrict processing, and to complain to a supervisory authority.

Deletion from the live service is immediate and irreversible. Precisely:

• Workspaces only you belong to are deleted outright, with every project, question, idea, report, saved opportunity, monitoring rule and note in them.
• Workspaces you share with other people are not deleted. Your membership is removed. The work in them belongs to that team and stays with them. If you were the owner, ownership passes to the longest-standing remaining member so the workspace is never left ownerless.
• Your identity, profile, sessions and acceptance records are deleted.
• Security event records are kept, with your identifying details removed from them. Deleting the record of security-relevant events along with the account would remove the only trail of what happened; the entry remains, you are no longer named in it.
• Your sign-in identity is held by our authentication provider and is removed separately by an operator. The application cannot delete it directly, and we would rather say so than imply otherwise.
• Encrypted backups made before your deletion still contain your data until they are replaced on the schedule below. They are not used to run the service and are only ever read to restore it after a failure.

Your account data is kept while your account exists, and goes from the live service when you delete it. Public source material collected by Discover is not tied to your account and is retained as part of the shared corpus. Minimised security records are retained for security purposes.

Backups are made daily. The most recent seven daily backups, four weekly backups and three monthly backups are kept and older ones are removed, so a copy of deleted data can remain in a backup for up to about three months.

Specific retention periods for security records and request logs have not yet been set. They are recorded as an open decision rather than stated as a number we do not actually enforce.

Access requires authentication, and every read is scoped to the workspaces you belong to. Session IP addresses are stored hashed rather than in the clear, invite tokens are stored hashed, requests are rate-limited, and backups are encrypted before they are stored. No system is perfectly secure, and we do not claim this one is.

This policy carries a version. If it changes materially you will be asked to review the new version before you continue using the service.