Nutrition logging, hypertrophy programming, and weight-trend analysis in one iOS app — built around a campus dining database so a dining-hall meal logs as precisely as a packaged food.
Nutrition apps are built around packaged food. Scan a barcode, get exact macros. But most of what a student on a meal plan eats never had a barcode, so the only options are guessing or logging something that is merely close. Do that three times a day and the numbers stop meaning anything.
CalCamp treats the dining hall as a first-class food source: menus are pulled live from Rutgers Nutrislice, so a campus meal is as precise an entry as a scanned box — and the rest of the app is built to hold that same standard.
Barcode scanning that resolves through FatSecret's GTIN-13 lookup, food search across USDA and Open Food Facts, on-device nutrition-label OCR, photo estimation, and live campus dining menus. Anything the camera reads can come from a photo taken earlier instead.
A 232-exercise catalogue, user-created exercises, routine templates with supersets, and a live session logger with rest timers, plate maths, Brzycki 1RM estimates, and personal records.
Weight-trend smoothing that separates real change from water and food, TDEE estimated from adherent days, and weekly volume per muscle with direct and assisting work counted separately. Thirty-five micronutrients are tracked against twenty targets, each with a plain-language note on what it actually does.
Apple Health read and write, Live Activities on the Lock Screen and Dynamic Island, campus geofencing, offline-first sync, and 13 languages including right-to-left layouts.
Macros are normalised per serving, and a nutrient the source never reported is stored as missing rather than collapsed to zero. A zero is a claim about the food; missing is the truth, and the difference compounds across a day of logs.
A set can be weight × reps, bodyweight reps, a duration, or a distance. Forcing every movement into one schema would make the training data quietly wrong, so the schema bends instead.
A custom biomechanical volume-tracking algorithm attributes assisting work to the muscles that actually did it. Counting every set at full value against every muscle involved is what makes naive trackers fire false overtraining alerts, so the model splits direct and assisting volume instead.
The Today screen says “2 saved on device · queued” instead of pretending the log is already synced. Knowing a meal is held locally and will go up later is the difference between trusting the tracker and logging the same lunch twice.