Your plan reacts to what happens in your body

5 systems, one goal: train exactly what you need

After each workout, we check how you are for tomorrow

When you complete a workout, the system calculates your TSB (Training Stress Balance) for the next day. If it falls below -10, an automatic analysis is triggered. If it drops below -25, it directly proposes replacing the session with recovery.

The AI looks at your accumulated fatigue, the trend over the last 5 days, what you did today, what you have planned, and the context of the week. With all that, it decides what to recommend.

The system has 4 response levels depending on severity:

If you accept the recommendation, the AI regenerates tomorrow's session incorporating the fatigue considerations. If you dismiss it, everything stays as it was. It never changes anything without your confirmation.

Your Garmin tells us how you slept and how your body is doing

When your Garmin sends HRV and sleep data for the same day, the system evaluates whether today's workout is compatible with your recovery status. You don't need to do anything: it arrives automatically.

The AI combines multiple signals to avoid overreacting to a single data point:

• ✓ HRV vs your average — Compares last night's value with your 7-day average and analyzes the trend (rising, falling, stable).
• ✓ Sleep quality — Score from 0 to 100, total duration, breakdown by type (deep, REM, light, awake).
• ✓ Signal convergence — One metric out of range does not trigger an alert if the other is fine. It only acts when there are multiple indicators.

If the weather is dangerous, the coach warns you before you go out

Every morning, the system checks the weather forecast for your location. If the feels-like temperature exceeds 32°C or drops below 5°C during training hours, an automatic analysis is triggered.

We know how you will feel before you even start

When you open a workout for today or tomorrow, the system automatically calculates whether the session will feel different from what was designed. It combines 6 factors into a single calculation:

• ✓ Accumulated fatigue (TSB) — The heaviest factor. From -0.5 (fresh) to +2.0 (critical fatigue).
• ✓ Consecutive days — The third day in a row does not feel the same as the first. +0.25 per extra day.
• ✓ Recent performance — Low scores on recent workouts = lingering fatigue.
• ✓ Sleep and HRV — Real-time Garmin data. Poor sleep or low HRV = everything feels harder.
• ✓ Your personal history — If you always report that workouts feel +2 above what was planned, the prediction calibrates to your real perception.

At the end of the week, the coach recalibrates what comes next

The system compares what was planned with what you actually did: sessions completed, total minutes, accumulated TSS, zone distribution, normalized power.

If the deviations are significant, the AI adjusts the following week automatically: it may reduce or increase minutes per day, change the focus of sessions, or recalibrate intensity. If everything went as planned, the following week stays the same.

This is the system that closes the loop: if during the week there were multiple adjustments for fatigue or wellness, the evaluation detects it and recalibrates to prevent it from happening again.

How the 5 systems connect

They do not work in isolation. They share data and reinforce each other:

A coach that does not wait until you get injured

• ✓ Proactive — Detects early signals (TSB dropping, HRV falling, sleep deteriorating) and proposes adjustments before there is a problem.
• ✓ Multiple data sources — TSB from training, HRV and sleep from Garmin, RPE from the user. Signal convergence reduces false positives.
• ✓ Conservative — Only acts on clear signals. Does not bombard you with alerts for normal fluctuations.
• ✓ Never without your permission — Every recommendation requires your confirmation. The system proposes, you decide.
• ✓ Works without Garmin — Without HRV or sleep data, it still works with TSB, consecutive days and RPE. Garmin enriches it, but is not required.
• ✓ Transparent — You see exactly what data was analyzed, what was found, and what is recommended. No black boxes.