Book a demoEach client gets their own context model.
For every client Harlen constructs a context model, their notes, check-ins, injury history, wearables, programming, and results, and a capable AI reasons over it. The longer you coach with it, the sharper it gets.
Book a demoIllustrative. Hover a node: notes, HRV, resting heart rate, injury history, preferences, programming.
The visualization above is an animated network of one athlete’s context: thousands of connected data points forming a single sphere, with denser hubs for coach notes, HRV, resting heart rate, injury history, preferences, and programming.
A context model per client
Compiled per client from notes, check-ins, injury history, wearables, programming, travel, and results.
A capable AI reasons over it
We do not build the AI. A leading model reasons over each client's context model, and every read sharpens as those models improve.
Every surface draws from it
The morning read, answers about any client, programs, and briefs all come out of the same context model.
Better the longer you use it
Every morning, check-in, and note adds to the record. It learns that specific person.
FAQ
Questions about the model.
Are you building an AI model?
No. The AI is a leading model we plug in, and swap as they get better. What we build is the context model: your whole structured record of each client, which is what the AI reasons over.
What is a context model built from?
About a dozen sources per client: notes, check-ins, injury history, wearable data, programming, training, travel, and results, all read against that person's baseline.
Does it diagnose or give medical advice?
No. It reports what changed against each athlete's own baseline, not a cause. Anything clinical stays with you and their physician.
Will it message my athletes or change a program?
Never on its own. Harlen surfaces what changed and suggests a next step. It does not contact your athlete or edit a plan. You make the call.
Does it get better over time?
Yes. Every check-in and note becomes part of that client's context model, and the reasoning improves as the underlying AI does too.
Is every client’s context model separate?
Yes. Each athlete gets their own context model. One client's data never colors another's read.
Can I trust the answers?
When it reasons about training or recovery, it draws on published research and cites what it used, so you can open the paper and check it yourself.