Race readiness, in one number.
A 0–100 score from fitness, fatigue, and adherence. Measured — never narrated by the model.
81 / 100
Strong · trending up
You’re putting in the miles. But are you actually on track for race day? Azimuth reads your runs, sleep, and recovery every morning — then gives you a clear signal, not a guess.
Strava tracks what you ran. Garmin tracks how you slept. But nothing tells you what it adds up to — or whether you’re actually on track for race day. That’s the gap Azimuth closes.
Not more data to scroll through. Five clear readings from your actual training, waiting for you at 6am.
A 0–100 score from fitness, fatigue, and adherence. Measured — never narrated by the model.
3am wake-up with the baby? Business travel? The block target holds. The path to it adjusts. You get the week you can actually run.
Threshold pace, easy HR, stride, zone time. Updated every session, never inflated.
One paragraph on what worked, what slipped, and the week ahead. Generated only when there's enough signal.
AI-writtenSync once. Every run flows into the plan, gets categorised, feeds the readings — and the model.
The narrated card explains the change. The measured card proves it. You’re not left wondering why your threshold session moved — you can see exactly what changed and why.
HRV trended low for three days after travel. I rebalanced this week's threshold into easy 12k and held the goal pace untouched.
Past weeks fade. Tap a row to open the session. The long-run column never lies — it's how race day will feel, mostly.
Threshold pace, easy HR, zone time, stride. Each updates from your runs — not a self-rated slider.
Every score is produced by deterministic code — it doesn't hallucinate, it doesn't round up. The model reads those numbers and explains them. It's a writer, not a decision-maker.
You can't score a tune-up race you haven't run. So Azimuth doesn't pretend you scored zero — components with no data drop out of both sides of the calculation. Your score is always paired with an evidence completeness percentage.
Fitness Level moves fast — it answers “how is training going right now?” Race Readiness moves slowly — it answers “how prepared am I for race day?” Showing both is more honest than blending them into a single number.
No. It's the read between the sessions a coach prescribes, or the structure for a self-coached runner. Many beta users use Azimuth alongside their coach as a daily check-in.
It writes three things — a daily read, a weekly review, and a plan-adjustment note. It never sets your readiness score, never sets your goal pace, and never decides your plan. It narrates.
It looks at the last seven days — sleep, HRV trend, completed sessions, perceived effort, weather — and rebalances the next three. The block target doesn't move; the path to it does.
No. Your training data stays yours. Inference happens on the seven-day rolling window required for the writeup; raw runs never leave Azimuth.
You can — millions of runners do. But a standard plan doesn't know you've had three nights of poor sleep. It doesn't know your HRV has been trending down for ten days. It doesn't know the threshold session you planned is the one that tips you toward injury. Azimuth knows — and adjusts before the damage is done.
Nothing during the beta. After Berlin '26 we'll publish a fair, single-tier subscription. Beta runners get founding-member pricing, locked.
Not hoping. Knowing. Drop your email — the next cohort opens March '26. Your first week: a readiness baseline, a goal-pace conversation, and a plan built around your actual life.