Your Training Data Context
Stride AI builds a picture of you from your history, current fitness, recovery and life outside training. That lets it set volume, intensity and progression for you rather than following a one-size-fits-all plan. For example, if you have trained fairly consistently on Tuesdays, Thursdays and Saturdays, the AI notices and builds a similar week.
Current metrics#
The AI starts with your current profile to set targets and limits.
- FTP / CP: used to scale workout intensities, set your training zones and estimate the load of each session.
- W′ (W prime): shapes high-intensity and repeated work, especially intervals above threshold.
- Threshold pace and heart rate: set run and swim targets and heart rate zones.
- Weight: used for watts per kilo, and to keep power targets realistic on climbs and in longer events.
Recovery and sleep#
- HRV, resting heart rate and breathing rate: the AI looks at the last seven days of each against your own 30-day baseline, not just today's number. A few days of falling HRV with rising resting heart rate is often the first sign of illness or accumulated fatigue, and it eases your sessions when it sees that pattern.
- Sleep and recovery scores: from whichever device you have chosen to lead your sleep and recovery (see Integrations).
Power curve#
Your power curve shows what you can do across durations, not just one FTP number. The AI uses it to understand your strengths and limiters (strong 5-minute power but weaker 30-minute power, say) and shapes intervals and race-specific work around them. It can also look at your durability: how much power you keep after hours of work.
Load and Fatigue#
Your Load and Fatigue tell the AI how much work you can handle and how fast to progress.
- Load: your long-term training load, built up over weeks and months, sets the baseline for how much more you can take on.
- Acute load and Fatigue: acute load is roughly your last week of training. Fatigue is Load minus acute load, so a higher number means you are fresher. The AI uses both to decide when to push, when to hold and when to plan an easier block.
Your full history#
The AI can look back to your very first recorded activity, not just the last few weeks. It can scan any period session by session and then open individual activities in full detail. Weekly summaries include how many sessions you did, your biggest session and your longest, so it can compare this year's build with last year's.
Past activities#
Past sessions show the AI how you respond to different kinds of work, not just how much you do.
- Training Score (TS): measures the impact of each session, so the AI can balance hard, moderate and easy days.
- RPE and feel: how hard a session felt and how you felt doing it. Rate it on the activity in Stride after each session. If your Garmin asks for RPE when you save, that comes through too.
- Comments: notes about sleep, stress, pain or how a session went, typed or left as a voice note (voice notes are transcribed, so the AI reads what you said). Your coach's comments are read too.
- Session type and structure: intervals, endurance rides, races and group rides are read differently when the AI looks at load and adaptation, and it checks how closely you hit each planned interval.
Strength and cycle tracking#
If you log strength sessions, the AI sees your sets, the weights you lift and the equipment you have, and uses them when it writes your next session. If you track your menstrual cycle, it sees your current phase and when your next period is expected, and adapts training to how you say you feel.
Illness, holidays and events#
Life context matters, so the AI does not treat you like a machine. When you set up your calendar with events and constraints, it uses them to shape your plan:
- Illness and injury: time off or reduced capacity is taken into account, so the AI rebuilds gradually instead of dropping you straight back into full training.
- Holidays: known breaks let the AI load up beforehand, ease off, or maintain around those dates.
- Events and races: their dates, priorities and your notes set the build and taper. For long events, previous finishes of the same event help it judge pacing.
- Availability and notes: an early meeting, a busy afternoon or any other constraint. You can leave any kind of note for the AI to read.
Memory and past conversations#
The AI keeps notes about your goals and preferences between conversations, and can search your earlier chats. See Training Planner for how to read and edit its memory.
Where the data comes from#
Stride AI uses everything synced to your Stride account, including data from Garmin. The limits on Garmin data described in Connecting Stride to Claude (MCP) apply only to outside AI assistants, not to Stride's own AI.