TroponinIQ and the 6-Week Recomp: Coach Judgment at Scale

TroponinIQ and the 6-Week Recomp: Coach Judgment at Scale

AI coaching is useful when it compresses recall, routine analysis, and pattern spotting—but the coach still owns the calls that matter.

Justin Harris told a client during a planned 6-week recomp that the client was “leaner” even though sick-related water retention was blunting the visual change. That is the mechanism in one phrase: pattern recognition plus contextual override. The sharp thesis is simple: AI can widen coach leverage across a roster, but only if it helps the coach remember, compare, and standardize without replacing the judgment that decides when a signal is noise. If it starts making the call instead of supporting the call, scale turns into shallow automation.

The bottleneck is not information

Most serious coaching problems are not caused by a shortage of data. They are caused by the wrong person or system carrying the burden of recall, comparison, and timing.

Justin’s own message to a client spelled out the real friction: “So many things are repeated topics, it’s hard to remember who I discussed them with and who I didn’t.” He added that he was “hyper paranoid” about seeming inattentive, so he repeated things “a dozen times just to be safe.” That is a coach-leverage problem, not a science problem. When the roster grows, the issue is no longer whether the coach knows the answer. It is whether the coach can reliably remember which athlete needs which answer, and when.

That is exactly where AI coaching tools can help, but only in a limited and practical way: not by inventing insight, but by reducing the time spent reconstructing context. For a coach managing multiple athletes, the value of AI is not mystical personalization. It is keeping track of the boring, repetitive, high-frequency stuff so the coach can reserve attention for exceptions.

Why the blood sugar note matters

In one client exchange, a post-show athlete described blood sugar issues as expected rather than alarming. The coach’s response was not to escalate emotionally or overcorrect. It was to recognize that the rebound context was already known, that the issue had been anticipated, and that the main job was not panic management.

That matters because it shows where leverage lives in coaching judgment. A tool can record that a client reported a blood sugar pattern, or that a high day required an insulin dose reduction, or that meals had to be brought closer together. But the coach still has to decide what is actually meaningful in context. In that same exchange, the client said the same insulin dose as the week before made the meal 1 blood sugar dip noticeably, so meal timing had to move up by 30 minutes. After reducing the dose by 1 IU, the same pattern still happened on the third shot, so the plan was to reduce further on the next high day.

That sequence is useful because it shows the shape of coaching work at high resolution: observe, compare to baseline, adjust in small steps, watch for repeat behavior, then adjust again. AI is useful here only if it helps surface that sequence quickly across many athletes. The coach still has to know that the problem is not merely “blood sugar,” but changing sensitivity in a specific diet phase with a specific response to a specific dose.

Scale is not the same as standardization

A lot of AI coaching hype quietly assumes that if a process can be standardized, then judgment can be removed. That is backward.

Justin’s nutrition guidance in the deep knowledge base is a good example. On carb sources, the message is not “fruit is always good” or “fruit is always bad.” The actual guidance is conditional: the vast majority of results come from nailing macros; pre- and post-workout fruit can cover up to about 50% of the carbs in those meals with no downside; on medium days fruit can be used in other meals as long as it is not every meal and at least half the carbs per meal come from more complex sources; on high days, when carbs are very high and insulin is elevated all day, sugar barely matters and fruit helps keep food volume down.

That is exactly the kind of rule set AI should be able to operationalize. It can surface the guideline, remind the coach what day type the athlete is on, and flag when a meal log drifts away from the rule. But the judgment remains human because the rule is not a slogan. It is a conditional map tied to day type, meal composition, food volume, and athlete adherence.

In practice, that means AI should be used to compress administrative load, not to flatten the decision tree. A coach with 20, 50, or 100 athletes does not need another chatbot that speaks in confident generalities. The coach needs a system that remembers which athlete is on which day type, what the last adjustment was, what pattern repeated, and where the exception is hiding.

What AI should actually do for coaches

The useful jobs are not glamorous:

• keep a clean record of prior adjustments so the coach does not have to rely on memory;
• summarize repeated issues across a roster so patterns become visible sooner;
• separate routine updates from true exceptions;
• preserve the chain of reasoning from observation to adjustment.

Those are leverage multipliers because they let one coach maintain quality while the roster grows. But there is a hard line: AI should not be the thing deciding whether a change is warranted. The coach decides that.

Why? Because the best coaching messages in the KB are not just correct; they are context-sensitive. The sick athlete on a recomp was told not to overread the pictures while water was still present. The blood sugar rebound athlete was told, in effect, that the pattern was expected and the concern level should stay controlled. The carb source guidance changed by day type. None of those are universal answers. They are judgments made from context.

The preservation test

Here is the falsifiable test for AI fitness coaching:

If the system makes a coach faster at finding the right context, but the coach still has to choose the adjustment, it improves leverage.

If the system makes the adjustment for the coach, it degrades judgment.

That distinction sounds small, but it is the whole game. Coaches do not scale by cloning themselves into software. They scale by turning repetitive memory work into machine work while keeping the actual call human.

So the real promise of TroponinIQ is not that it replaces the coach’s eye. It is that it gives the coach a better memory, a cleaner roster view, and less admin drag, so judgment stays available for the moments when context matters most.

Sources Used

• raw/_consumed/2026-05-31/kahunas-export/2026-05-31-w13-18m/clients/joe_webb___members-rksigkykimaxwmo_t4_e8nwvbtc2j0etleutkyysads.json
• raw/_consumed/2026-05-31/kahunas-export/2026-05-31-w13-18m/transcripts/joe_webb___members-rksigkykimaxwmo_t4_e8nwvbtc2j0etleutkyysads.md
• raw/_consumed/2026-05-31/kahunas-export/2026-05-31-w19-24m/clients/rory_lazowski___members-c5balaovjbdoeefqmfuqdhh2tbpmfdu16lnf0tnrtmw.json
• raw/_consumed/2026-05-31/kahunas-export/2026-05-31-w19-24m/transcripts/rory_lazowski___members-c5balaovjbdoeefqmfuqdhh2tbpmfdu16lnf0tnrtmw.md
• modules/03-knowledge/kahunas-coaching-deep-nutrition.md