Names & Faces
I work at a theater conservatory. A few times a year a brand-new cohort walks through the door, and somebody hands me a PDF — pages of headshots with names printed underneath. Wonderful people. Total strangers. And there is nothing warmer on this earth than someone who's met you exactly once saying "Morning, Coraline" like it was nothing.
So: you import the PDF. That's the whole workflow. In a couple of seconds, entirely on your phone, the app renders every page, finds every face, reads the name printed under it, and deals you a deck of flashcards. No typing 23 names. No cropping 23 photos.
The game is the part I actually care about. You start with three faces — nobody learns a whole cohort at once, and the app doesn't pretend you can. When you're feeling dangerous, you add one. New faces come around often; faces you keep nailing earn their way down to rare. Blank on someone you "knew" and their frequency quietly climbs right back up. You never manage any of this. The deck just pays attention.
Vision and PDFKit do the extraction on-device — no servers, no accounts, no "we value your privacy" theater. The students' photos never leave the phone. Cost to run: $0.00, forever. And if you don't have a PDF handy, it deals you a starter deck of US Presidents so you can finally learn which one is Chester A. Arthur.