Product entry point
The website explains the connection between Intervals.icu, STAS, ChatGPT, and Claude.
Static asset: /images/search-growth/intervals-ai-coach-home.png
Press and reviewer kit
STAS.run connects Intervals.icu training data to ChatGPT and Claude
STAS is a practical AI coaching layer for endurance athletes who already collect training data in Intervals.icu. It helps ChatGPT or Claude answer from real workouts, load, fatigue, calendar context, goals, and saved rules.
Try STAS with Intervals.icu dataSee product screenshotsShort positioning
STAS is not a new watch app and not a closed training-plan generator. It sits between Intervals.icu and AI chat, preparing the data an athlete would otherwise have to copy, summarize, or screenshot by hand.
- For runners, cyclists, triathletes, and mixed-load endurance athletes.
- Best for athletes who already use Intervals.icu or are willing to connect it once.
- Useful when the athlete wants explanations and plan discussion, not only an automatic workout.
- Works with ChatGPT and Claude as the coaching conversation layer.
How the data flow works
The product uses Intervals.icu as the source of truth, then gives AI enough context to discuss training decisions in normal language.
Watch or app
Garmin, Suunto, Polar, COROS, Amazfit, Apple Watch, and other sources can sync training into Intervals.icu.
Intervals.icu
Workouts, calendar, load, fatigue, form, zones, wellness, and history stay in the athlete's Intervals.icu account.
STAS
STAS prepares recent workouts, current condition, sport context, goals, rules, and plan context for AI use.
ChatGPT or Claude
The athlete asks for review, explanations, safer alternatives, weekly planning, or plan rewrites.
Who STAS is for
The strongest early fit is an endurance athlete who has real training history and wants AI to reason from that history instead of giving generic advice.
Intervals.icu users
Athletes who already use Intervals.icu for analysis, fitness/fatigue tracking, and calendar planning.
Self-coached athletes
People who want help reviewing training, but still want to understand and approve plan changes.
Athletes with imperfect weeks
Useful when travel, stress, missed workouts, soreness, strength work, or poor sleep changes the plan.
Reviewers and writers
A compact example of AI coaching that uses real training data rather than a standalone chatbot prompt.
What STAS can help answer
These are the kinds of questions STAS is built to support once Intervals.icu is connected.
Current condition
Is this athlete fresh, overloaded, rebuilding, or ready for intensity this week?
Workout review
Was the session really easy, too hard, or different from the planned intent?
Plan adjustment
How should the next 7-10 days change after a missed session, fatigue spike, or race priority?
Training explanation
Why is this workout recommended, and what is the safer alternative if recovery looks poor?
What STAS actually does with Intervals.icu data
STAS does not train a separate personal model from private data. It fetches and prepares the athlete's current training context so ChatGPT or Claude can reason from it during the session.
Reads training history
Recent workouts, sport type, duration, intensity, pace, power, heart rate, and structured workout details where available.
Reads load signals
Fitness, fatigue, form, weekly history, calendar context, zones, and wellness or recovery signals when Intervals.icu has them.
Prepares coaching context
Profile, goals, rules, current condition, strategy, and sport-specific summaries are shaped into context the AI can use.
Supports plan work
The AI can discuss a plan rewrite and, when the user chooses, STAS can help write planned sessions back to Intervals.icu.
Screenshots reviewers can use for product context
These screenshots are already available as public assets and show the real product flow rather than mock launch messaging.
Workout context
Training details show the kind of real workout data that supports coaching answers.
Static asset: /images/search-growth/intervals-ai-coach-training-detail.png
AI coaching context
The AI conversation can start from prepared training context instead of pasted screenshots.
Static asset: /images/search-growth/chatgpt-real-training-context.png
Connection setup
The setup path is centered on connecting Intervals.icu data and then using an AI channel.
Static asset: /images/search-growth/chatgpt-onboarding-connections.png
Simple reviewer checklist
1
Open the product story
Start with the four linked pages below: they explain the Intervals.icu data flow, ChatGPT use case, Amazfit HR-zone checks, and Zepp Coach context.
2
Check the screenshots
Use the public proof assets on this page for a quick visual understanding of setup, training data, and AI coaching context.
3
Connect data for a real test
The practical test is connecting Intervals.icu and asking about current condition, a workout review, or a short plan rewrite.
4
Note the limits
STAS supports training decisions. It is not medical advice, not emergency guidance, and not a substitute for a human coach when the situation requires one.
Useful facts for a concise review
Not a watch replacement
STAS depends on existing training data sources and Intervals.icu rather than replacing the athlete's watch ecosystem.
Not a generic chatbot prompt
The difference is data access: recent workouts, load, zones, calendar, goals, and rules can be available in the conversation.
Not full automation
Plan changes are intentionally discussable. The athlete can ask why, choose a safer option, and approve what changes.
Privacy and limitations
STAS needs permission to read Intervals.icu data so it can prepare useful coaching context. The athlete remains in control of account connection and training decisions.
- Intervals.icu remains the source of truth for training history and calendar data.
- AI output can be wrong; athletes should use judgment and avoid training through injury or illness.
- Medical, injury, nutrition, and high-risk advice should come from qualified professionals.
Reviewer notes
Is STAS a training-plan app?
Not primarily. STAS is a data bridge and coaching context layer for ChatGPT and Claude. It can help discuss or write plans, but the main value is AI seeing real Intervals.icu data.
Does it require Intervals.icu?
Yes for the core product promise. Intervals.icu is the data layer that keeps workouts, load, zones, calendar, and history together.
Which sports are relevant?
The product is strongest for endurance workflows: running, cycling, triathlon, and mixed weeks that include strength or other activity.
What should a reviewer test first?
Connect Intervals.icu data and ask about current condition, one recent workout, and whether the next week should change.
Four P0 pages for deeper review
AI coach for Intervals.icu
The main explanation of STAS as an AI coaching layer on top of Intervals.icu.
ChatGPT with Intervals.icu data
A practical page showing why ChatGPT needs real workout and plan context.
Amazfit HR zones troubleshooting
A concrete proof path around heart-rate zones, Intervals.icu history, and STAS interpretation.
Zepp Coach guide
A balanced guide on Zepp Coach, load checks, and when to review broader training history.
Try STAS with real training context
Connect Intervals.icu data, then ask STAS through ChatGPT or Claude about a recent workout, current condition, or the next week.
Try STAS with Intervals.icu dataFree STAS account. Intervals.icu connection is required for the real data workflow.