The coach problem
Most coaching teams already have data: training logs, wearable exports, testing results, readiness scores, and notes. The hard part is deciding what, if anything, is worth changing for one athlete this week.
A Causal Driver Audit is built for that narrow decision. It works from exported athlete data and returns a coach-readable packet that says what the system found, how strong it is, and where the evidence stops.
What AthDash returns
- Driver Cards that name the driver, outcome, estimated effect, interval, sample context, and license to advise or withhold.
- Evidence states that distinguish supported, exploratory, borrowed, and insufficient relationships.
- Caveats and next-data notes so the coach can see what would make the finding stronger or unusable.
- Agent-readable evidence packets for AI coaching workflows that need deterministic evidence before drafting advice.
What it does not claim
AthDash does not diagnose, treat, or prevent injury or illness. It does not guarantee performance changes. It does not replace a coach, clinician, or safety judgment. A relationship marked insufficient is returned as a refusal state, not softened into advice.
Exports first
Start with CSVs, logs, and dashboard exports before promising integrations.
Driver Cards
Readable findings with evidence, interval, caveat, and wording boundary attached.
No-claim states
WITHHOLD and DECLINE are valid outputs when the evidence is too thin.