SlopeReady exists because of a problem we had ourselves. We ski, and we wanted to show up to the start of the season actually fit for it rather than spending the first two weeks getting hammered on every run. So we started training earlier in the year. And almost immediately we hit the same wall every motivated amateur hits: how do you know if any of it is working? How do you know a given week is moving you toward the goal instead of just making you tired?
Our first instinct was the obvious one in 2026 — ask an AI assistant like ChatGPT. Paste in some numbers, describe how the week went, ask what to do next. Sometimes the answer was genuinely good. But using it that way, day after day, turned out to be a chore, and eventually a little misleading.
What actually went wrong when we used AI chat
The first thing was the friction. Your health data does not come as tidy text you can copy. It lives in Apple Health and in your watch, spread across screens. So you end up screenshotting a sleep graph, typing out "resting heart rate felt high this morning, legs a bit heavy from Tuesday," and then trying to work out what the model wants to hear before it will say anything useful. Every session started with ten minutes of transcribing yourself.
The second thing was worse. An AI assistant like ChatGPT can retain a conversation, a memory, or a project, and that can be useful. But it does not automatically become a current, inspectable coaching record tied to the data that actually reached Apple Health, the workout you moved, and the plan you are following. In our own use, that gap let the plan drift. Ask on Monday and you got one program; ask again a few days later, phrased slightly differently, and you got a different one. We saw noticeably different reads on the same day depending on how we described things that morning versus that afternoon. None of it was wrong exactly — it was just untethered. Nothing held it to what we had actually been doing.
That is the gap we set out to close. Not "AI assistants are bad" — they are genuinely useful reasoning tools, and putting a training question into words is already half the work. The problem is that a one-off answer to a pasted fragment is not the same thing as coaching, and it is easy to mistake one for the other.
An AI chat sees the snapshot you paste, not the picture
When you paste last night's sleep and this morning's heart rate into an AI chat such as ChatGPT, it can reason only from the context that conversation has. It will not automatically know your resting heart rate has crept up over the last five days, that your load has been climbing for a fortnight, or that today's run was slower than your usual easy pace at the same effort.
Those comparisons are the part that changes the answer. A single high number is rarely the story. The story is how today sits against your recent normal — and to reconstruct that in an AI chat you would have to maintain the history, keep it current, and already know which comparisons matter before you know what you are looking for.
SlopeReady works from the other end. It reads the context you have already built up in Apple Health — recent workouts, sleep, resting heart rate, HRV when it is there, training load, recovery signals — so a question about today gets answered against your actual recent weeks instead of a fragment you retyped from memory.
Coaching remembers the plan you are following
An AI assistant like ChatGPT may retain parts of a conversation, but it does not automatically own a durable plan state. Ask it the same thing next Tuesday and it may not have a reliable record of the plan it suggested, the workout you moved, and the goal you set. You become the person keeping that record aligned — which is exactly how it becomes easy to end up with a different program every few days.
SlopeReady keeps a plan that is yours and stays put between sessions. When life gets in the way you can ask to move a workout, shorten it, or swap it for something easier, and the plan adjusts while protecting what mattered about the week. The goal is not to preserve every box on the original calendar. It is to keep the goal intact when the schedule has to bend.
Coaching knows when to hold you back
An AI assistant can hand you an ambitious plan whenever you ask, even when the input is too thin to judge whether it fits your recovery. Coaching sometimes has to say "not today" and mean it for a reason you can see.
SlopeReady leans on explicit limits rather than enthusiasm. When recent load is high relative to what you are used to, or recovery signals say you are run down, it eases the intensity and volume instead of pushing. An easy or recovery-focused day is a normal, deliberate output here — not the app losing its nerve.
It should be willing to say "I do not know"
Health data is not a clean lab record. Maybe you did not wear the watch overnight, HRV did not sync, or a workout showed up without enough detail to compare against your usual sessions. With a partial snapshot, any answer can sound more certain than the input deserves unless it makes the gap explicit.
We built SlopeReady to do the opposite. Missing data stays missing rather than getting invented, and readiness carries a sense of how complete the underlying picture actually is. When the evidence is thin, the honest move is to keep the day conservative and not draw a big conclusion from one odd reading. That is also why we explain how to start when your data is incomplete. SlopeReady is a training tool — not a doctor, and not a replacement for a coach or clinician who knows you personally.
You should be able to see what it sees
When you paste data into an AI assistant such as ChatGPT, it is hard to know afterwards what it actually used, what it assumed, and where that information ended up. For anything touching health, that is not a small detail.
SlopeReady is built so raw Apple Health data stays on your device, and when you use an AI feature it works from a derived, minimized, consent-gated view of your context rather than your whole health record. You can inspect exactly what the coach was given, so an answer is something you can check instead of something you take on faith.
Built around skiing, useful the rest of the year
Skiing is our main hobby, so that is what SlopeReady is tuned for — arriving at the season with legs and lungs that can handle it, and staying durable once you are on snow. But the same machinery works for the rest of your training. It is happy being a day-to-day companion for running, cycling, strength, or just keeping your general condition honest through the off months. You keep recording wherever you already record — Apple Watch, Strava, Garmin, Fitness, whatever you like — and SlopeReady reads what reaches Apple Health and analyzes it alongside your recovery and your plan. Access is read-only, so it never starts a session, edits your history, or writes anything back.
So ask — but ask something that has the context
Talking to an AI assistant like ChatGPT to think out loud about training is fine; we still do it. The limit is not the AI model. It is that a pasted snapshot cannot carry your changing data, your plan, your safety limits, or an honest account of what is missing — and without those, the answer quietly drifts. Coaching is the part that holds all of that between conversations, so what you hear on a Wednesday fits the week you are actually in. That is the whole reason we built it.