Privacy Policy

What we collect

• Search queries — queries you submit in the Explore or Benchmark tabs are sent to the backend to produce results. They are not stored permanently or associated with your identity.
• IP address — used server-side to enforce per-IP rate limits (Explore AI answers and Benchmark runs). Not logged permanently.
• Access request email— if you submit the "Request access" form on the Benchmark page, your email, optional name, and optional note are stored in AWS S3 and used only to send you an access key. We do not share this data with third parties or use it for marketing.
• Uploaded documents — if you use the advanced benchmark feature to upload your own files, they are stored temporarily in AWS S3 and automatically deleted after 7 days. They are not used for training or shared.

What we do not collect

We do not use cookies, analytics trackers, or advertising SDKs. We do not create persistent user accounts or user profiles.

Third-party services

• Amazon Bedrock — LLM-only and RAG queries are processed by AWS Bedrock (Claude Haiku). Queries are subject to AWS Service Terms.
• GitHub Pages— the static frontend is hosted on GitHub Pages and subject to GitHub's privacy policy.

Data deletion

To request deletion of your access request email, contact us via GitHub at github.com/Raxorr/ragornot. We will remove it within 30 days.

Changes

This policy may be updated at any time. Material changes will be noted in the project repository.