Software that adjusts loads from RIR
Your client reports RIR when the set ends. Kaizer recalculates the estimated 1RM and writes the next load into the weeks you did not touch. The ones you touch stop.
Short answer.
Kaizer adjusts loads from the RIR your client reports on each set. When they finish the session, the platform takes the logged sets, reps and RIR, recalculates the estimated 1RM using NSCA-derived tables, and writes the weights and reps for the upcoming sessions. It only does that on the weeks you have not edited: the moment you touch one, that week drops out of the calculation permanently. And you decide who it runs for, because automatic progression switches on and off globally and client by client. We call it assisted progression: the arithmetic gets automated, the judgment does not.
Updated: August 3, 2026
See Kaizer with your own use case.
When this flow actually changes something.
You have more clients than you can recalculate by hand each week, and progression is the first thing that slips.
You program with fixed percentages of 1RM and you know a client's bad day never shows up in a spreadsheet.
You already ask for RIR or RPE and the number ends up in a WhatsApp message nobody reads again.
You want the load to move on its own where you cannot get to it, and to stop where you already decided.
Clients plateau, and when you open the routine you find nobody moved a variable in six weeks.
How it works, step by step.
The client reports RIR
At the end of each exercise the app asks how many reps were left. You can turn that question on or off for all your clients or client by client.
The estimated 1RM is recalculated
Kaizer combines the weight lifted, the reps achieved and the reported RIR, and estimates the 1RM with NSCA-derived tables. No maximal test required.
The next load is written
With the updated 1RM, the platform computes the weights and reps for that client's upcoming sessions of that exercise.
Your edit wins
The engine only writes weeks nobody touched: the moment you edit the weight or the reps of a week, that week drops out of the calculation permanently. And automatic progression switches on and off globally and client by client. That is what makes it assisted rather than automatic.
The whole routine as one table
Open the full routine as a table and edit reps, weight and RIR week by week, without going into each session.
The history stays
Every session records whether it was trained, when, at what loads and at what RIR. Each exercise's progression shows on the client's profile.
Frequently asked questions
Kaizer does. It takes the sets, reps and RIR the client logged in the app, estimates the 1RM using NSCA-derived tables and uses that value to compute the next session's loads. The recalculation runs after every logged session, not once per mesocycle.
The mechanism is direct. If the prescribed set finished with more reps in reserve than expected, the real effort was lower than planned and the load goes up. If it went to failure or fell short of the rep target, it comes down. Kaizer applies that reasoning per exercise and per client, on the weeks you have not edited yet.
Yes, and that is the point. What gets automated is the arithmetic: reading the logs, estimating the 1RM and computing the next load. What does not get automated is the decision, and you have three controls to exercise it: a generated routine does not reach the client until you send it, any week you edit is frozen against the engine, and automatic progression can be switched off globally or for one client.
Most progress on a fixed percentage of 1RM or on increments set in advance, without looking at what happened in the session. Kaizer starts from the real effort reported on each set, and stops where you already decided: the week you edit leaves the calculation for good. That matters because the log does not capture what you do know: that they slept four hours, that they are coming back from an injury, that form broke down on the last set.
That is the real risk of the scale, especially with beginners, who systematically underestimate how many reps were left. That is why the RIR question can be turned off per client, and so can automatic progression: if you know a client under-reports, you switch it off for them or write the week yourself, and the engine never touches it again.
No. The 1RM is estimated from submaximal working sets, which is how it works in practice: almost nobody tests a true max every week because it is exhausting and it carries risk.
Load adaptation is available from the Free plan at USD 0, up to 3 clients. Paid plans start at USD 19 per month and scale by client count.
