How it works
Each session teaches it more about you, so the next plan is better.
Try it.
A small, tappable version of the app. Example data only.
Lower body
40 minutes. Gym.
No overhead pressing today. Shoulder flagged.
0 of 9 sets done
Tap a set to log it.
Today
Targets follow your weight trend.
Tap a meal to add it.
How are you?
Flag anything that hurts.
Nothing flagged. You get the full plan.
Your week
Sunday review. Example.
One thing next. Add 2.5 kg to your back squat on Monday.
Example data. The real app uses yours.
Connect your data
Sleep, heart rate, steps and bodyweight come in from Apple Health each day, so the plan reflects how you actually are. If an upload is missed, the next one fills the gap.
Set your goals
Pick what you're after: a lift, a body composition change, a race time, an injury to get past, or just consistency. Goals shape every session and every weekly review.
Choose today's time and place
The app first decides the type of session from what you've done recently and what's sore, with no AI involved. Then it asks how long you have and where you are.
Get a session
The AI picks exercises for that time and equipment, applies rehab work if you have an active injury, and suggests loads. Sleep, recovery and nutrition adjust how hard it pushes.
Train, log, note how it felt
Log sets, add a note on any exercise, and flag pain at the end. Notes feed into the next session's plan.
Nutrition adjusts to the trend
Your calorie and macro targets are corrected by the gap between the weight change expected and the one observed, after accounting for lifestyle context.
Review the week
Every Sunday evening a short review lands with what moved, what stalled and what to focus on.
What the AI does
It chooses exercises, writes the coaching and explains why. It reads a structured summary of your data, not a raw dump.
What the code does
It works out the numbers: trend weight, expenditure, calorie and macro targets, load suggestions and goal progress. The AI is handed those values and never does arithmetic.