Who is responsible

[Legal entity name], [Postal address], is responsible for your data. Write to [contact email] about anything on this page.

What we collect

• Your email address, to send sign-in links and receipts.
• What you give the agent: uploaded papers, pasted paper lists, your question, any details you add, and a company website address if you start from one.
• What the agent produces: the passages it extracts, the claims it verifies, the list of sources it looked at, and your report.
• Payment records from Stripe: what you bought, when, and whether it was paid or refunded. Stripe processes your card; we never see or store the card number.
• Technical data: a cookie that keeps you signed in and ties your investigations to you, and the request logs our hosting provider keeps (IP address, time, page).

We use no advertising or analytics cookies, and we do not sell your data. We use your papers and questions to produce your reports and run the service, and for nothing else.

Who processes your papers

The agent runs on AI models from these providers. Each receives only what its part of the work needs, through its API:

• Anthropic (Claude) runs the agent itself and reads your papers to extract and check claims. It receives your papers, your question and the sources the agent reads. The agent also uses Anthropic's web search. Anthropic's privacy policy.
• OpenAI grades the agent's draft report independently, so the agent is not the only judge of its own work. It sees the draft report and the agent's working notes, which can include excerpts of your papers and your question. OpenAI's privacy policy.
• OpenRouter carries the screening step, which decides which candidate sources are worth reading. It receives your question and the titles and abstracts of public candidate sources. Every request we send through OpenRouter is restricted to model providers that do not collect data for training and to endpoints with zero data retention; if no such endpoint is available, the request fails rather than going anywhere else, and screening falls back to Anthropic. OpenRouter's privacy policy.

To find related sources, the agent sends search queries it writes from your question and papers to search services: Exa, Tavily, OpenAlex and arXiv. It then fetches openly available pages and papers from the sites that publish them. If you start from a company's website, it reads that site's public pages.

Where it is stored, and for how long

• Uploaded files are stored privately in Google Cloud Storage in the United States (us-central1) and deleted automatically 90 days after upload.
• Your account, investigations and reports, including the passages extracted from your papers, are stored in a database run by Neon in the United States, and kept until you ask us to delete them.
• The service runs on Google Cloud in the United States.
• Sign-in emails are sent by [Email delivery provider].

Who can see your reports

Your investigations are private to you. The published investigations on this site are ones we ran ourselves; we do not publish yours without your permission.

Your rights

You can ask for a copy of your data, for a correction, or for your account and investigations to be deleted: write to [contact email] from the address you sign in with. Depending on where you live, you may also complain to your data protection authority.

Changes

If we change this policy, the effective date above changes, and we will tell you by email before a change that affects data we already hold. See also our terms.