Lassor won Profound’s applied Agentic AI for Marketing demo competition.
Lifestyle & fitness

Swiper Fitness

The in-gym fitness logging app that allows trainers to build routines and track progress with their clients.

  • SwiftUI
  • React
  • Supabase
  • Vercel
  • Resend

I built Swiper from the ground up, including the iPhone and Apple Watch apps, database architecture, hosting, web design, branding, copy, and marketing.

The opportunity

Personal training has moved online — more than half of trainers coach clients over video. But few software offerings accommodate this use case.

Swiper lets trainers and clients collaboratively build routines, log workouts, and track every set. With an unmatched interface that pairs simplicity and elegance, it’s the highest-quality in-gym logging experience for clients and the best way for trainers to track progress.

TrainerizeTrueCoachMy PT HubPT DistinctionJEFITFitbodHevyStrongSwiper
Coaching platforms are sprawling all-in-one suites, and the best focused loggers are built for people training alone. Nothing gives trainers and clients a fast, focused log to share. Competitive positioning map. Coaching platforms built as all-in-one suites: Trainerize, TrueCoach, My PT Hub, and PT Distinction. All-in-one apps for individuals: JEFIT and Fitbod. Focused logging apps for individuals: Hevy and Strong. Swiper is alone in the corner for focused tools built for coaches.

The product

Swiper is the simplest way to log a workout: every set is a single swipe, on iPhone or Apple Watch, with nothing to learn and nothing to slow you down.

And it’s just as simple to share. Invite your trainer in, and they can build your routines and see every set you log, so you’re always working from the same plan, even when you’re training alone.

The interface is built to disappear. Each set asks for one swipe and nothing more, so lifters stay in their flow between sets instead of in their phone, and the next thing to do is always the only thing on screen. Animated demo of a Swiper workout on an iPhone and an Apple Watch showing the same exercise. Three bench press sets are swiped complete on the phone, and the list scrolls to cable fly. The first set is tapped open on the phone and its weight raised from 40 to 45 pounds while the watch shows the same edit, then the set is swiped complete on the watch. The watch crown scrolls to plank, where two 30 second bodyweight timed sets count down from three and run their clocks before completing.

The system

Making the trainer-client relationship simple took as much work as making the gym experience focused and clear. The two workflows share one architecture, built directly into the interface.

A client can invite their trainer in to plan their workouts and see how each one went. The trainer writes the plan, the client follows it swipe for swipe, and both always see the same picture: what was planned, and what actually happened.

SWIPER · DATA MODELTrain alone, coached togetherEvery set your trainer plans and every set you log live in the same place.TrainerCoaches over videoClientTrains in the gymSHARES ACCOUNTClient lets trainer· build routines· start workouts· review historyBUILD TOGETHERRoutineA reusable plannameownerHAS MANYExerciseIn the planexerciseorderHAS MANYPlanned setThe targetreps, weighttype: reps | timedWorkoutOne session of a routinestarted, active timefinished_atHAS MANYExerciseAs performedexercisecopied_fromHAS MANYLogged setWhat actually happenedreps, weightstatus, logged_atSTARTSCOPIED TOCOPIED TOFinished?YESHistoryEvery finished workoutprogress over timeREVIEWED BY TRAINERpersonplannedperformeddecisionlinked copy
One model serves both sides: it keeps the solo workout fast and focused, and makes sharing with a trainer genuinely useful, because every set they plan and every set you log live in the same place. Diagram of Swiper’s data model. A trainer and a client share an account; the client grants permission to build routines, start workouts, and review history, and the two build routines together. A routine has many exercises, and each exercise has many planned sets with target reps and weight. A routine starts a workout, which has many exercises, each with many logged sets; each workout exercise and logged set is copied from, and links back to, its counterpart in the routine. A decision asks whether the workout is finished; if so it goes to history, which tracks progress over time and is reviewed by the trainer.

The outcomes

I built Swiper to teach myself to build software with AI, and it’s grown into something people rely on. Today it’s used almost every day by a small community of trainers, their clients, and me.

The iPhone and Apple Watch apps are native SwiftUI, and the web app is React and TypeScript, built with Vite. Behind them is a stack I now use for most of my projects: Supabase for the database, authentication, and real-time sync, and Vercel for hosting and deployment. Trainers and clients share one data model, so a routine a trainer builds shows up instantly on a client’s iPhone or Apple Watch, and every set they log flows back the same way.

The graph below tracks that activity. Every square is a day from the last year, and the brighter it is, the more workouts were finished.

973 workouts in past year
Last updated 11:06 PM PST
A contribution graph, the grid of daily squares GitHub shows for each programmer’s work, repurposed to count every workout Swiper’s users finish.