How SuppleSense reads your data

SuppleSense surfaces patterns in your own log using a deterministic statistical method, not opaque prediction. For each candidate relationship — a supplement and a subjective rating, or a supplement and a biometric — we compute a Spearman rank correlation over the days you have logged, alongside the sample size that underwrites it.

We display a result only when the sample size is at least fourteen days and the false-discovery-controlled p-value clears a calm threshold. The denominator and the confidence interval are shown on every insight. A small sample stays a small sample — we do not present a noisy result as a strong claim.

Large-language-model narrative is used to describe what was observed, never to invent a relationship the statistics did not already earn. The model can choose how to phrase "associated with"; it cannot decide what is associated.

Not medical advice

SuppleSense is a personal data tool. Patterns are observations from your log, not clinical guidance. Talk with a qualified clinician before changing supplements, peptides, hormones, or prescription medications.