How Scatter handles your data.

Terms of use

Scatter is provided as-is for informational purposes only. Patent family data is sourced from the European Patent Office (EPO) INPADOC database; prosecution data is sourced from the patent offices and public file-wrapper services listed under Data sources below. While we strive for accuracy, this tool does not constitute legal advice. Always verify critical data against official patent office records before relying on it for legal or business decisions.

AI-assisted extraction (examination report analysis, citation extraction) is experimental. Results should be reviewed by a qualified patent professional before use in prosecution, opposition, or litigation contexts.

Accounts and personal information

Scatter offers optional accounts. Browsing and searching do not require one. If you create an account we store your email address and a session record so you can stay signed in. Anything you do while signed in is stored against the account as well: documents you upload, the families on your watchlist, and whether you want update emails. Sign-in links are delivered by our email provider, Resend. You can ask for your account and its data to be deleted at any time via the feedback form in the app.

Recent searches and pinned families are stored in your browser, not on our servers. A search query reaches our servers only so that patent data can be fetched from the patent offices. We remove application numbers from the page addresses sent to our analytics, but a number can still reach analytics in the text of something you click or in the details of a browser error report.

Cookies

A session cookie keeps you signed in if you have an account. We use Vercel Analytics and PostHog for usage analytics (page views, feature usage) and to report errors that happen in your browser. Session recording is off. Analytics are not tied to your account email.

Document processing

When you use the prosecution analysis feature, uploaded PDF documents are processed server-side to extract structured data. Document content is sent to OpenAI for extraction. A scanned document is sent to Microsoft Azure for text recognition before that. Extracted text and metadata are stored in our database to avoid re-processing on subsequent visits. A document you upload while signed in is private to your account and is not shown to other users. We keep the original file so you can view it in full and so it does not have to be processed again.

Sub-processor data retention terms and our processing and storage regions are being finalised; this page will state them before open account signup is announced.

Data sources

• Patent family and bibliographic data: EPO Open Patent Services (INPADOC)
• AU prosecution history: IP Australia AusPat
• EP prosecution history: USPTO Global Dossier, with the European Patent Register as a fallback
• US prosecution history: USPTO Open Data Portal (Patent Center)
• NZ prosecution history: IPONZ
• JP, KR, CN and CA prosecution history: USPTO Global Dossier
• BR prosecution history: INPI Brazil
• AR prosecution history: INPI Argentina
• FI prosecution history: PRH Finland
• PCT/WO prosecution history: USPTO Global Dossier, with the AU and EP file wrappers of national-phase descendants as a fallback
• Cited document PDFs: EPO Open Patent Services, with the USPTO's own document download as a fallback for US publications
• English translations of cited documents: the issuing office where it publishes one, otherwise Google Patents or machine translation by Scatter
• Citation extraction: AI-assisted (configurable provider)

Material from IP Australia: Licensed from the Commonwealth of Australia under a Creative Commons Attribution 4.0 International Licence (creativecommons.org/licenses/by/4.0). Scatter has modified it: records are extracted, reformatted and combined with data from other sources.

Contact

For questions, feedback, or account deletion requests, use the feedback widget in the app.