From notebook to check-in: how AIron runs your clients
Most coaches already have their program — in a notebook, a spreadsheet, their head. The work starts after that: every week you check who trained, watch the videos, compare what was done against the plan and decide next week's loads. Here are two steps AIron does for you — shown with the app's real components.
1. A program from a notebook photo
Pick the client, open the AI chat and attach a photo of the page. The AI reads days, sets, rep ranges and progression rules, and you get an editable proposal. Only when you apply it do the sessions land on the client's calendar.
UPPER / LOWER — 4 days
Upper A: bench 4×6-8, row 4×8, OHP 3×10
Lower A: squat 4×6, RDL 3×8, lunge 3×10
Upper B / Lower B — same, swap bench↔OHP
Plan
4 days · 12 exercises
The example is a classic 4-day upper/lower split with linear progression: top of the rep range → +2.5 kg (upper) or +5 kg (lower), two misses → −10%.
2. The check-in: the week in one place
Open Clients → the client → Calendar. Above the calendar sits the check-in bar: the latest review in one tap, or "Check in now" to review this week. The demo below is the same card you use in the app, with sample data.
Strong week: 3/4 sessions, squat ready for +5 kg
3/4 planned done · extra: 1
- Missed sessions
- Pain mentioned
- New videos
- Squat and bench: every planned rep completed.
- Lower B was missed — rest of the week on track.
- Row: lower-back pain → hold.
Past week · W39
Wed 30Lower BMissed
Thu 1Upper BDone5 km walkExtra
- Barbell rowplan 3×8 · 60 kgdone 3×7 · 60 kg
To review · 2
- Back squatMon, 28 Sept
- Barbell rowThu, 1 OctPain mentioned
Last set was a grind, slight lower-back pain.
Next week · Sat 3 – Sat 10
to change: 2 · unchanged: 3Sat 3 · Upper A
Sun 4 · Lower A
Tue 6 · Upper B
Message to the client (draft)
1 · Tap “Check-in” in the client's header — a draft only you can see is prepared. Nothing is sent.
What a check-in contains:
- Adherence — planned done / planned due. Unplanned sessions count separately ("+1 extra") and never inflate the percentage.
- Needs your eyes — the client's videos and notes. A mention of pain is flagged, and that lift gets no increase.
- Plan vs. performed per lift across the last two exposures.
- Next week — a suggestion (increase / add a rep / hold / deload) with the reason. "Apply" updates the planned sessions of the next 7 days — once, so a change can't be doubled.
- A draft message in your language. You edit and send it yourself — nothing reaches the client automatically.
You can also schedule check-ins (e.g. Sunday 18:00) or run them from ChatGPT/Claude through the MCP connection — every one lands in the same history.
More things that speed up client work
- Sidebar priorities — who is running out of program, who hasn't trained in X days.
- Copy / paste a week in the calendar, with progression.
- Exposure sheet — every lift's history in a grid, with targets for future sessions.
- Team programs — one program shared by coaches and assigned to clients.
What this demo covers, and what it doesn't (honestly)
- The check-in card is the real component; the data is sample data and the demo's buttons are disabled.
- The numbers (adherence, plan vs. performed, suggestions) are computed deterministically. The AI only writes the headline, the summary and the draft message.
- Check-in loads follow the client's check-in settings (default +2.5 / +5 kg, −10% deload). Rules stored in the program itself are not yet carried into the check-in automatically — if your program uses different steps, set them in the check-in rule.
- Reading the photo is an AI model: review the proposal before applying, especially with messy handwriting.
Try it on your next workout
Start free