baseline
Own your baseline.
Know what is actually working. Every stream of your health data in one record — then Bayesian causal inference on your own history. You get an effect size with a credible interval, not a correlation and a vibe.
How Baseline works
Baseline merges six streams of personal health data — wearable and watch metrics, blood panels, gut and continuous glucose monitoring, meals, symptoms, and supplements — into a single record aligned on one nightly time axis. It then fits a hierarchical Bayesian time series to that record, pooling across a population to estimate the causal effect of an intervention on an outcome for one individual.
Worked result
Magnesium glycinate 400 mg increased deep sleep by 18(95% credible interval +6 to +31 minutes) across 62with a 97% posterior probability of benefit. Estimated with a hierarchical Bayesian model, n=1 pooled across the cohort, holding travel, alcohol, and late meals as confounders.
N1Bench
The first open benchmark for causal reasoning over personal health data. We publish our own scores next to every frontier model — a system that tells you what’s working should be able to prove it doesn’t tell you things that aren’t.