Auto edge detection with live camera preview
Real-time capture makes scans feel fast and native instead of clumsy and delayed.
iOS App Development
Native iOS app. Zero to App Store in 3 days.
Papyr is a privacy-first document scanner for iPhone - no ads, no subscriptions, no cloud uploads. Built with AI-assisted development and shipped in a single weekend.
- 72 hours
- zero to App Store
- 13,000+
- LOC of production Swift/SwiftUI
- 210
- requirements verified across 4 audits
- 0
- third-party dependencies
The Challenge
Build a scanner that does not turn private documents into someone else's data.
The brief was straightforward: build a document scanner that respects user privacy.
Every option on the App Store was either ad-infested, subscription-locked, or quietly uploading documents to someone else's cloud.
Papyr needed to feel first-class on iPhone while keeping scanning, OCR, search, and export fully on-device from day one.
What the operation was fighting
- Ad-heavy experiences that interrupt a simple utility workflow
- Subscriptions gating basic scanning and export behavior
- Cloud uploads where sensitive documents leave the device
What We Built
A focused native scanner with the essentials done properly.
Papyr shipped as a real product, not a weekend prototype.
Full OCR with searchable text across documents
Every scan becomes searchable locally, without sending pages through a cloud OCR pipeline.
PDF export with optional password protection
Users can generate shareable PDFs and lock them down when documents need extra protection.
Folder organization and batch scanning (Pro)
Pro workflows support multi-page capture and cleaner structure for people scanning more than one sheet at a time.
Dark mode, accessibility, and full VisionKit integration
The app was built to feel at home on iPhone, including native platform behavior and accessible defaults.
How We Built It
Native code, zero dependencies, and a tight spec-to-audit loop.
Papyr was built in 72 hours, from an empty Xcode project to App Store approval.
The stack is native Swift and SwiftUI on iOS 17+, with VisionKit, Vision, Core Image, PDFKit, SwiftData, and StoreKit 2 handling the heavy lifting.
Everything runs locally: scanning, OCR, search, PDF export, and document management. There are no network requests anywhere in the codebase.
AI-assisted development made the timeline possible. Claude handled architecture decisions and spec writing. Codex handled implementation. A rigorous spec-to-audit-to-ship pipeline verified 210 requirements across 4 complete audits before release.
Implementation notes
- Native Swift/SwiftUI, iOS 17+
- VisionKit, Vision, Core Image, PDFKit, SwiftData, StoreKit 2
- AI-assisted: Claude for architecture, Codex for implementation
- 4 complete spec audits, 210 requirements checked
The Result
A privacy-first iPhone app shipped in a single weekend.
Papyr went from zero to the App Store in 3 days and landed with Apple's privacy label set to Data Not Collected.
The result is 13,000+ lines of production Swift and SwiftUI, a complete on-device document workflow, and an app that never sends a single document over the network.
Papyr gives iPhone users a scanner without ads, subscriptions, or surprise cloud behavior - exactly what the brief called for.
- 13,000+
- lines of production Swift/SwiftUI
- 210
- requirements verified across 4 audits
- Data Not Collected
- Apple privacy label on the App Store
- 0
- network requests in the codebase
Next Step
Start Your Conversation
If you need a product shipped quickly without cutting the hard engineering corners, let's talk.
hello@rowvyn.com