The Signal Score
The Signal Score is fully deterministic. Physical and Cognitive sub-scores combine into a single 0–100 number and a Red / Amber / Green traffic light — the same inputs always produce the same result.
Composite
66
Cognitive reaction time trending down pulled the composite into Amber. The LLM only narrates this — it never alters the number.
Each domain contributes to either the Physical or Cognitive sub-score. Every signal is traceable back to the device that produced it.
Resting heart rate, HRV, and recovery curves.
RHR · HRV · VO₂ est.
Architecture, latency, and consistency over time.
Deep · REM · efficiency
Movement load balanced against recovery capacity.
Steps · load · strain
Reaction time and sustained-attention drift.
RT · attention · variance
Autonomic balance and stress load trends.
SNS/PNS · recovery
Weight, body fat, and longitudinal drift.
Mass · BF% · hydration
Rings, watches, bands, and scales all flow through one ingestion pipeline — including Hume Health, with an optional Hume 0–900 scale view for teams that prefer it.
Connect a deviceAsk it anything about your health in plain language. It reads your own numbers, explains what they mean, and remembers what matters to you between conversations — so the advice gets more useful over time, not more generic.
Your goals, your routine, and the interventions that actually worked for you — carried from one conversation to the next, so you never start from scratch.
View, edit, or erase anything it remembers, any time. Your memory is yours — it’s there to help, not to lock you in.
It puts your deterministic results into plain language. It cannot move your score, change a traffic light, or raise an alarm — those are computed, not generated.

The number that matters is fixed math. The AI’s job is to make it make sense — and to do the quiet work in the background.
The Signal Score is a fixed, published formula. Same inputs, same number — reproducible and independently checkable. The AI cannot change it.
Where we infer a trend or a biological age, it comes with a stated confidence band that updates as more data arrives. No false precision.
It turns the numbers into plain language. It cannot move your score, change a traffic light, or raise an alarm — those are computed, never generated.
Photograph a meal and the analysis runs on your phone — the image never goes to the cloud.
A background research job keeps the list of supported wearables and scales up to date on its own.
For HR teams
Every HR view enforces a minimum cohort size. Groups too small to anonymize are suppressed entirely.
Calibrated noise is added to aggregate metrics so no individual can be reverse-engineered from the totals.
Employees own their data. HR sees only DP aggregates. Signal Labs Admin staff never browse individual rows.