Platform
See exactly what you get.
Below are replicas of the app's own screens, filled with real output for a sample member: a training session the AI wrote and audited, a meal the food log read, and the maths behind the member's targets. The photo report is an example, because real ones are private.
01 · Training session
Written for one person. Checked before you lift.
This is a real session the app wrote for a test member, laid out the way a member sees it when they open the day. It is research-grounded AI programming: written from a summary of published muscle-building research and the member's own logged sets, for one person, then checked by code against fixed rules. Everything the agents wrote at the top of the session is shown in full.
Thursday · Week 2 of 4
Light Legs (Quads) & Upper Body + Core
16 movements · 48 sets · est ~77 min
✓ Meets your requirements: 13 lifts (min 12) · 3 abs (min 3) · weekly minimums on track
Coach
Annotation: ProgrammerJordan, tomorrow is your conditioning class, so today we focus exclusively on building the quads and upper body while keeping your hamstrings and calves completely fresh. You’ve hit a few rep plateaus over the last month, so today is about breaking them. We are forcing concrete load jumps on your hack squat, incline DB press, shoulder press, and weighted core movements. Put the heavier weight in your hands, keep the execution ruthless, and let's force the body to adapt.
Why this session:
Annotation: ProgrammerProtects the posterior chain for tomorrow's conditioning by capping quads at 7 sets and glutes at 4, with zero hamstring, calf, or lower back work. To address the heavy stagnation list, mandatory load jumps were applied to every applicable movement (hack squat, Bulgarian split squats, incline DB press, underhand pulldowns, reverse pec deck, and core work). Stagnant calf work was omitted strictly due to the conditioning constraint and will be hit next leg day. Ab wheel rollout was swapped for a weighted plank to break the anti-extension plateau.
Photo check-in:
Annotation: Programmer · photo analysisAnnotation: ExampleWeakest area first: the lower abs and obliques still show little separation, and that has not changed since your earliest photos. From the side, the upper chest is flat compared with the lower chest. Side delts are wider than in the first set, but still trail your traps. Back thickness and rear delts have clearly responded. Lighting differs between the oldest and newest front shots, so treat small changes in definition with caution.
This block is illustrative: the sample member had no photos on file, so the real session has none. With photos, it sits here at the top of every session.
Audit ✓ (deep review)
Annotation: AuditorAdjusted the draft to ensure strict adherence to the conditioning class limits (no hamstrings/calves/hinge, quads capped). Increased leg extension and straight-arm pulldown to force progression since they hit the top of rep ranges. Swapped Ab Wheel Rollout (stalled) for Weighted Plank. Included Machine Shoulder Press and Incline DB Curls to actively force jumps on Stagnation Watch items rather than omitting them. Corrected DB Bulgarian Split Squat tags to [glutes, quads]. Ensured 13 main and 3 ab exercises with all rest periods and supersets configured correctly.
Written by AI from your logs. Not medical advice — stop if anything hurts.
Warm-up
- 5 minutes easy on the stationary bike
- Dynamic leg swings and walking lunges (bodyweight)
- Band pull-aparts and arm circles
- 1 set of light machine hack squats to groove the pattern
Pressing Enter on a blank set logs the target. That counts as done, but not as proof you can go heavier — type what you actually lifted.
Machine Hack Squat
Target: 4 × 8-10 @ 190 lbs · Rest 2 min · quads, glutes
Coach: Last logged 180x8. Forced jump to 190 lbs to break the plateau. Keep tension continuous through the quads.
Dumbbell Bulgarian Split Squat
Target: 2 × 6-8 @ 60 lb DBs · Rest 2 min · glutes, quads
Coach: Last logged 55x6. Jumping to 60s today to force overload. Deep stretch at the bottom.
Leg Extension
Target: 3 × 10-12 @ 200 lbs · Rest 0s · quads
Coach: Superset with Hip Adduction. You hit 195x12 across all sets last time; moving up to 200.
Machine Hip Adduction
Target: 3 × 10-12 @ Find it: a weight you could lift for 3 more reps · Rest 90s · adductors
Coach: Superset with Leg Extension. Smooth and controlled.
The other 12 movements, condensed for this page
In the app each one is a full card like those above, with its own coach's note.
- Incline Dumbbell Press3 × 6-8 @ 70 lb DBs
- Underhand Lat Pulldown3 × 8-10 @ 165 lbs
- Machine Shoulder Press3 × 8-10 @ 155 lbs
- Chest-Supported Machine Row3 × 8-10 @ 185 lbs
- Reverse Pec Deck Fly3 × 10-12 @ 125 lbs
- Machine Lateral Raise3 × 10-12 @ 95 lbs
- Incline Dumbbell Bicep Curl3 × 6-8 @ 45 lb DBs
- Overhead Cable Triceps Extension3 × 10-12 @ 140 lbs
- Straight-Arm Cable Pulldown3 × 10-12 @ 135 lbs
- Weighted Hanging Leg RaiseCORE3 × 8-12 @ 10 lb DB
- Decline Weighted CrunchCORE3 × 10-12 @ 50 lbs
- Weighted PlankCORE3 × 60s @ 55 lb plate
Completed: 0 · Remaining: 16
▶ Start Workout — Begin Timer02 · Nutrition
A meal in one sentence. Targets with the working shown.
Food decides whether the training pays off, so logging it has to take seconds. Here is one breakfast going in, the day it lands in, and how the scale steers the targets.
01
Type it the way you'd say it
- Type a meal in a sentence, or photograph the plate. The Meal Reader splits it into items, each with calories, protein, carbs, fat and fibre.
- The portions it assumed are spelled out before anything is saved. If they're wrong, discard it and say it differently.
- Foods you've eaten recently come back as one-tap buttons, so a usual breakfast takes one tap.
Nutrition
Type it the way you'd say it. Your coach works out the numbers.
Check this looks right
Annotation: Meal ReaderAssumed 'shake' means 1 standard scoop of whey protein powder mixed with 1 cup of 2% milk, and oatmeal is a standard 1 cup cooked serving; eggs include a small amount of cooking fat.
What the Meal Reader sent back for that sentence
Annotation: Meal Reader| Food | kcal | Protein | Carbs | Fat | Fibre |
|---|---|---|---|---|---|
| Eggs (cooked)3 large | 229 | 18 | 1 | 17 | 0 |
| Oatmeal1 cup cooked | 155 | 5 | 27 | 3 | 4 |
| Banana1 medium | 104 | 1 | 25 | 0 | 3 |
| Protein shake with milk1 scoop whey + 1 cup 2% milk | 242 | 32 | 15 | 6 | 0 |
| Meal total | 730 | 56 | 68 | 26 | 7 |
Eggs (cooked)
3 large
- kcal
- 229
- Protein
- 18
- Carbs
- 1
- Fat
- 17
- Fibre
- 0
Oatmeal
1 cup cooked
- kcal
- 155
- Protein
- 5
- Carbs
- 27
- Fat
- 3
- Fibre
- 4
Banana
1 medium
- kcal
- 104
- Protein
- 1
- Carbs
- 25
- Fat
- 0
- Fibre
- 3
Protein shake with milk
1 scoop whey + 1 cup 2% milk
- kcal
- 242
- Protein
- 32
- Carbs
- 15
- Fat
- 6
- Fibre
- 0
Meal total
- kcal
- 730
- Protein
- 56
- Carbs
- 68
- Fat
- 26
- Fibre
- 7
Grams, except kcal. Like any food log, these are estimates.
02
Targets with the working shown
- Your calorie, protein, carb, fat and fibre targets come from your bodyweight, height, age, sex, goal and training days. Plain code does this maths, not AI, and every line of the reasoning is on the page.
- Protein is set at 2.2 g per kg of bodyweight, fat at 0.8 g/kg as a floor, and carbs take the rest. Fibre scales with calories.
- Confirmed meals fill the day's bars as you go.
730/ 3030 kcal
2300 kcal to go
Eggs (cooked)
3 large
Oatmeal
1 cup cooked
Banana
1 medium
Protein shake with milk
1 scoop whey + 1 cup 2% milk
Where these targets come from
Annotation: Code · not AI- Maintenance estimated at 2793 kcal: BMR 1802 kcal from the male Mifflin-St Jeor equation, scaled for 4 training days.
- Muscle gain: +237 kcal a day (8% over maintenance), so 3030 kcal — about +0.47 lb per week. Faster than ~0.5 lb per week on a trained lifter is mostly fat.
- Protein 190 g is 2.2 g/kg — comfortably above the ~1.6 g/kg where muscle gains plateau.
- Fat 70 g is a floor for hormone health; carbs take the rest and fuel your sessions.
- Expect about +0.47 lb per week on the scale, judged on weekly averages, not single weigh-ins.
03
The scale steers your calories
- Weigh in most mornings. Single readings swing with water and food, so the app compares 7-day averages and measures the trend against the rate your goal planned for.
- Nothing is suggested until there are at least 8 weigh-ins to go on.
- Here the average moved +1 lb in a week against a planned +0.47 lb, so the dashboard says to pull back.
Morning weigh-ins, 21 Sep to 4 Oct
- Daily weigh-in
- 7-day average: 188.6 → 189.6 lb
All 14 readings
- 21 Sep · 188.4 lb
- 22 Sep · 187.9 lb
- 23 Sep · 188.4 lb
- 24 Sep · 188.9 lb
- 25 Sep · 188.3 lb
- 26 Sep · 188.9 lb
- 27 Sep · 189.4 lb
- 28 Sep · 188.8 lb
- 29 Sep · 189.4 lb
- 30 Sep · 189.8 lb
- 1 Oct · 189.3 lb
- 2 Oct · 189.9 lb
- 3 Oct · 190.3 lb
- 4 Oct · 189.8 lb
What needs attention
Annotation: Code · not AIGaining fast: +1 lb this week
Above roughly half a pound a week, most of the extra is fat rather than muscle.
→ Pull back about 200 kcal a day
This week
189.6 lb
7-day average · +1 lb this week
03 · AI photo analysis
Find what's lagging. Train it first.
Muscles don't all grow at the same rate. Progress photos show where you're behind, and the AI turns that into what your next sessions work on.
- 1
Upload front, back or any angle
Add progress photos whenever you like, from History & Photos. iPhone HEIC photos are converted in your browser before they're saved.
- 2
The AI compares, then writes
It sets your older photos against your newest, up to your six most recent. The report leads with your weakest area, says plainly what hasn't changed, and tells you when lighting or angle make the comparison unreliable. Honest, not flattering.
- 3
Your sessions aim at it
Weak spots are ranked worst first, three to five of them. Each session targets the top one that fits that day's muscles, and the exercise picked for it says so in its coach's note.
Oldest
Newest
Placeholders, not photos. The AI reads up to your six most recent, oldest to newest.
How to read a report
Leads with the weakest area
“Weakest area first: the lower abs and obliques”
Names what hasn't changed
“has not changed since your earliest photos”
Says when photos can't be compared
“Lighting differs between the oldest and newest front shots”
Ranks what to work on
“1. Core thickness”
Photo check-in:
Annotation: Programmer · photo analysisAnnotation: ExampleWeakest area first: the lower abs and obliques still show little separation, and that has not changed since your earliest photos. From the side, the upper chest is flat compared with the lower chest. Side delts are wider than in the first set, but still trail your traps. Back thickness and rear delts have clearly responded. Lighting differs between the oldest and newest front shots, so treat small changes in definition with caution.
Your photos, your call
Optional
Skip photos and there is simply no photo report. Training, nutrition and everything else work the same.
Visible only to you
Your photos appear only in your own account, and you can delete any of them at any time.
Analyzed by Google Vertex AI
To write your report, photos are sent to Google Vertex AI. Under Google Cloud's terms that data is not used to train Google's models. Details are in the privacy policy.
Can be switched off at intake
If your answers to the intake health check show that food or weight is a sensitive subject, photos are never sent for analysis and no photo commentary is written.
04 · Dashboard and trends
Both halves on one screen.
Training and food meet on the dashboard, and the trends page shows whether the work is actually increasing, not just whether you turned up.
Dashboard
- Muscles still under their weekly minimum, with sets done against the floor
- Lifts that have stopped moving, and the load they're stuck at
- How often you hit your protein, and your daily calorie average this week
- Your 7-day average bodyweight and how it moved
- Time per set, rest included
Four-week trends
- Total work per week: sessions, sets and tonnage, beside calories, protein and bodyweight
- Work per muscle, week by week
- Estimated one-rep max per lift, with stalled and climbing lifts listed
- Deload weeks shown but left out of every trend, so a planned easy week never reads as a decline
See it with your own numbers.
Everything above is built the same way for every member: sessions from your own logs, targets from your own body and, if you add them, reports from your own photos. One membership covers all of it: your first 7 days free, then $19.99 CAD a month or $199 a year, plus tax.
Apply or get invited. Every member is approved by hand, usually within a day. Your first 7 days are free, then $19.99 CAD a month or $199 a year, plus tax. Everything included, cancel anytime.