DealSnap

After every patient visit, our coordinator has to copy the handwritten sheet into our patient database by hand. It takes a couple of minutes each time, she does it 15 to 20 times a day, and it's easy to type the wrong name or file the patient under the wrong category.

I built a phone and tablet app that turns a photo of the patient sheet into a finished record. The coordinator takes a picture, AI reads the handwriting, figures out which treatment category and stage it belongs in, finds the matching patient, and creates the record. The whole thing takes about 30 seconds instead of a few minutes.

1. The coordinator unlocks the app with a PIN and photographs the patient sheet (plus an optional receipt for payment details).
2. AI reads the photo and pulls out the name, the treatment discussed, the dollar amount, and any payment made.
3. It sorts the patient into the right one of nine treatment categories and the right stage, and detects whether a deposit was paid.
4. It looks the patient up in our database and double-checks against our medical-records system so it links to the right person.
5. The coordinator sees a pre-filled form, fixes anything that looks off, and taps once to save.

A task that used to take two to three minutes per patient now takes about 30 seconds, cutting roughly an hour of daily typing down to about ten minutes. Every record lands in the right place, linked to the right patient.