Inksignal vs JunkDrawer

Reading on a screen can feel slow and tiring. Your eyes jump line to line, your focus breaks, and long reading lists keep growing. Inksignal gives you another way to read. It uses RSVP (Rapid Serial Visual Presentation): words appear one at a time at a fixed focal point, so you can keep your place and move faster with less eye strain. Built for students, researchers, and knowledge workers. What you can do with Inksignal: - Read at your pace: 100 to 1000+ words per minute with live speed control - Import real content: TXT, Markdown, PDF, and EPUB - Paste any URL and read extracted article text - Pull text from screenshots and photos with OCR - Load YouTube transcripts from video links - Control playback from the keyboard (play, pause, jump, speed up/down) - Stay focused with a minimal interface that gets out of the way Use it to review class material, clear article backlogs, or process docs before meetings. If you read a lot and want to do it faster without burning out your eyes, Inksignal is built for you.

Every iPhone has a thousand screenshots. The parking spot at the airport, the OTP from the bank, the auto driver's address, the kid's exam result, the prescription you snapped before the chemist closed. Photos already saves them. Finding them, three months later, is the problem. JunkDrawer indexes every screenshot you share to it - fully on your device - so you can find anything later. By word, by ₹ amount, or just by the vibe of what it was. WHAT IT DOES • OCR over English. Every word inside the screenshot becomes searchable. • Apple Intelligence classifies each screenshot - receipt, prescription, ticket, contact, invitation, social post, exam result, food, map, document, and more. • A one-line summary so you can skim a hundred screenshots without opening any. • Structured extraction. Phone numbers, OTPs, ₹ amounts, dates, addresses, UPI handles, IFSC codes, dosages, tracking numbers - pulled out as tappable pills. One tap to call, copy OTP, open Maps, copy the total. SEARCH THAT ACTUALLY WORKS Full-text + semantic + visual labels, all at once. Type "biryani" or "₹440" or "diabetes meds" - the right screenshot surfaces. No mode switching. No syntax. DOMAIN-AWARE • ₹ formats in lakhs and crores when they should. • UPI handles aren't treated as broken email addresses. • IFSC codes are first-class. • Phone numbers in +91 format pretty-print correctly. FULLY ON-DEVICE This matters more for screenshots than for almost any other category of personal data. Your screenshots are your bank app, your prescriptions, your OTPs. • No account, no cloud, no analytics, no telemetry. • OCR runs on Apple's Vision framework. Classification runs on Apple Foundation Models. Embeddings come from Apple's Natural Language framework. Nothing leaves your phone. • Delete the app and every screenshot it indexed is deleted with it. Your Photos library is never touched. • If iCloud Backup is on, your drawer rides to your next iPhone automatically. • Manual export creates a single .junkdrawer file you can stash anywhere — iCloud Drive, AirDrop to your Mac, email yourself. Tap one to import on a fresh install. HONEST ABOUT STATE • Pending screenshots show a small dot until they're analyzed. No fake completeness. • Settings tells you exactly which features are running and which aren't. • If the AI gets a kind wrong, change it from the screenshot's menu. Edit a wrong capture date the same way. WHAT IT ISN'T • Not another Photos app - your library stays where it is. • Not cloud sync (the schema is ready; sync ships in a later update). • Not a notes or journaling app. • Just one thing, done well. REQUIREMENTS Apple Intelligence required for full features (iPhone 15 Pro or later, iOS 26 recommended). Without it, full-text search and entity extraction still work on every screenshot - just no AI summaries or classification. Free forever, with an optional tip jar inside. Throw it in. Find it later.