Local PDF conversion
PDF text and page images are processed in the browser. Workers, fonts and OCR language assets are downloaded from the website origin. Documents and passwords remain browser state; reset discards them.
Open the guideLast updated: 2026-04-01
We process account data, authentication data, app usage counters, and note/group content required to provide Pocket Orbit features.
If you sign in with Google or Apple, we also process the identity data those providers return to us for account access, such as email address and provider account identifiers.
If you use AI summaries, your submitted note/group content is processed through our AI provider to generate summaries.
If you use OCR scan features, images you submit are processed to extract text and are handled transiently for conversion.
If you use the feedback board, we process public idea posts, votes, moderation reports, author-hide preferences, and private bug reports. Public ideas may be visible to other signed-in users, while private bug reports and moderation reports remain internal.
We use product analytics to understand feature usage, reliability, and high-level engagement patterns inside the app and landing experience.
We also process security and abuse-prevention data such as IP-based rate-limit signals and a hashed device fingerprint generated from device characteristics to help detect fraud, excessive multi-account creation, and misuse of free or trial limits.
OCR text extraction uses Google Cloud Vision as a processor.
Source images are processed for text extraction and are not stored as part of your account content unless you explicitly save extracted text as a note.
For the waitlist, we process your email address and anti-abuse verification data (reCAPTCHA token). Waitlist emails are kept until launch and then deleted after 2 months.
Main backend hosting: Hetzner Online GmbH (Germany). Landing frontend hosting: Vercel.
Depending on feature use, processors may include OpenAI (AI summaries), PostHog (product analytics), Apple and Google (Sign-In), Google (reCAPTCHA, waitlist sheet, Google Cloud Vision OCR), and Proton (support email).
You can request access, correction, deletion, restriction, objection, and data portability under GDPR/DSGVO. Contact: alpaca-labs.dev@proton.me.
Your local workspace
PDF text and page images are processed in the browser. Workers, fonts and OCR language assets are downloaded from the website origin. Documents and passwords remain browser state; reset discards them.
Open the guideMarkdown files remain in your workspace. Local AI processes selected material on your device. Choose a supported model in Settings and check each result against its source.
Open the guideModel downloads, store billing, optional accounts and referrals can require network access. Local processing does not promise that every app function works without a network.
Open the guideThe browser inspector processes text in the browser, makes no AI calls and sends no document content. Reference lists do not confirm whether destination files exist.
Open the guide