The signal underneath your biology.
Sleep, recovery, glucose, training, food — every vendor, every year, in one continuous record.
The ring sees one signal. Your phone sees another. The CGM sees a third. Continuum is what reads them together.
Good morning, Neil.
I'm still learning your patterns — think of me as a new teammate with a long way to go. Three readings I noticed overnight, or ask me anything.
6
first-class vendors
+ 400 via Terra-shape adapters
23
standardized fields
across sleep, HRV, glucose, activity
<800
ms median inference
Claude Sonnet 4.6 + cache
0
raw data egress
inferences leave; data stays
100
% signed responses
provenance attached, audit-ready
One inference layer. Four surfaces. Pitch the payer; force the players to integrate.
Continuum is the platform; the chat is just one consumer of it. Insurance, employer wellness, and pharma are the GTM wedge — concentrated buyers with millions of users each, and the leverage that forces hardware vendors to open their data.
Employers · payers · wellness platforms
Cohort insights without ingesting raw biometrics
Continuum sits between the wearable and the wellness platform. Aetna, BlueCross, Fortune 500 self-insured wellness, large employer benefits — they get burnout signals, recovery cohorts, and longitudinal outcome attribution as signed inferences. The biometric data never lands on their balance sheet. HIPAA-clean. BIPA-clean. GC-approved on the first call.
Pitch the payer; the players follow. When 10M employees access Continuum through their employer's wellness benefit, hardware vendors open data to stay reimbursable.
Pharma · supplement brands
Outcome attribution
Magnesium glycinate moved sleep score 8% in 4 weeks. GLP-1 preserved muscle mass vs control. Signed, vendor-agnostic, audit-ready. Pay-per-protocol pricing.
Clinics · longevity practices
Clinician extender
Continuous biometric monitoring + signed clinical context. Delivered into the EMR, not into a parallel dashboard. The 80/20 of telehealth follow-ups, automated.
Consumer apps · wearables
Drop-in inference for any wellness app
Oura's ring app, Whoop's recovery tab, a generic wellness coach — all consume the same Continuum API. The chat hero on this page is one such consumer. Partners ship faster; users get cross-vendor synthesis their hardware can't see.
Three layers. The middle one is the moat.
Continuum sits above the hardware. Wearables become peripherals; the inference layer is the product. Hardware vendors can copy the chat, but they can't copy a personal model that's been training across vendors for years.
01
Ingest
Any source, any vendor
Apple Health, Oura, Whoop, Dexcom, MyFitnessPal, Strava — connected via API. Export upload as the resilience path when APIs restrict. Apple can change their rules; the platform doesn't die.
sources
6
vendors
6
records / day
~12K
02
Synthesize
What no single vendor sees
Cross-vendor signal alignment, agreement scoring, lifetime baseline carried across hardware churn. Your 9pm glucose spike (Dexcom) caused your 2am HRV crash (Oura). No ring vendor can ship that.
agreement
94%
fields aligned
23
lifetime baseline
across vendors
03
Infer
The insight is the API
Every insight emits a provenance trace — sources, confidence, cross-vendor agreement. Downstream surfaces (Copilot, Slack, supplement brands, employer wellness) consume inferences, not raw biometrics. The Stripe pattern, applied to biology.
model
Sonnet 4.6
every response
signed
raw data leaves
never
Four things this is, that aggregator-apps aren't.
Bevel reads HealthKit and charges $6/month. That's a feature. This is a structural play — four architectural commitments that are hard for hardware vendors and platform giants to copy without dismantling their own business models.
Sensor fusion across vendors
Glucose × HRV × sleep architecture × calendar. Cross-modal correlations no single hardware vendor can see, because they only own one signal.
Lifetime baseline across hardware churn
Apple Watch → Whoop → Oura → next thing. The user accumulates 5–10 vendor models over a decade. Continuum is the only persistent record.
BYOD as the only viable enterprise motion
No Fortune 500 buys 50,000 rings. They adopt a platform that aggregates whatever is already on people's wrists. Agnosticism is the wedge.
Signed inference, not raw data
Every response carries provenance. Downstream platforms get the insight; the data never leaves. Employer wellness, supplement validation, clinical adjacency — all clear legal review.
The chat is the demo. The API is the product.
Three endpoints, signed responses, OAuth + Shortcut + export ingestion. Built for app developers, employer wellness platforms, and clinical-context partners who need biological inference without owning hardware or holding raw data.
// signed inference, one call
POST api.continuum.health/v1/infer
{
"user": "usr_…",
"query": "why am i tired today?",
"sources": ["auto"]
}
// → { insight, citations[], signature }