FAIT
Build a workout. Log every set.
Plan the next workout.

01Every set, logged.
A recorded demo session and its logged sets.
Bench Press
- 10 × 110 lbs
- 10 × 110 lbs
- 10 × 110 lbs
Deadlift
- 10 × 200 lbs
- 10 × 200 lbs
Pull Up
- 10 × bodyweight
Total lifted0 lbs
02Recovery, modeled.
Sets add estimated fatigue. Time reduces it. The app uses that estimate to plan the next split. A programmed estimate, not measured recovery.
- Chest46%
- Shoulders46%
- Triceps46%
- Back34%
- Biceps100%
- Abs100%
- Quads100%
- Hamstrings100%
- Glutes100%
- Lower back100%
Next workoutLower
- Push0.46
- Pull0.54
- Legs1.00
- Upper0.50
- Lower1.00
- Full body0.79
03A coach that writes workouts.
Profile, history and estimated recovery become context for a generated workout. Recorded example.
You ask
my chest is still a little sore, can you reccomend a good workout for today
The app adds
- Profile & goal
- BMI · RMR · TDEE · macros
- Last 5 workouts
- Recovery % per muscle
- Exercise catalog
The model answers OpenAI, through a Node proxy
Since your chest is sore, let’s focus on a lower body workout today. This will help you stay active without straining your chest muscles.
[{"name": …, "sets": …, "reps": …}]Parsed into a workout
- Squats3 × 12
- Lunges3 × 10
- Leg Press3 × 10
- Calf Raises3 × 15
- Plank3 × 1
My part
I worked on navigation, Isar data models, workout building and logging, recovery calculations, the body dashboard, and AI integration.
Team project with David Rodriguez and JReiniel.
Shows
- Modeling
- Exponential recovery modelRecovery-aware planning
- Apps
- Cross-platform Flutter appOn-device database (Isar)AI coach via API proxy
- Interface
- Data visualization
Built with
FlutterDart