Every wearable reads one signal. We read all of them — plus your DNA, your environment, and the latest science — without ever seeing your data.
VitaCrypt unifies your genetics, microbiome, wearables, labs, and environment into a single living health profile — encrypted on your device under a key only you hold, analyzed on a blind server that never sees plaintext. Targeting the Zama Concrete ML / TFHE stack.
The most valuable signals about your health — your DNA, your microbiome, your bloodwork — are the ones you should trust the fewest companies with. Today the deal is binary: hand over plaintext biology, or live with advice tuned to the median human who doesn't exist.
Fully Homomorphic Encryption changes the bargain. Your phone encrypts under a key only you hold (iOS Keychain / Android Keystore). Our compute servers run analysis directly on the ciphertext — mathematically indistinguishable from noise without your key. Apple shipped this same primitive (BFV) on every iOS 18 device. The cryptography just became consumer-grade.
Encryption is one pillar; the other is the scope of what we unify. Wearable APIs (unified via Terra API), open environmental data, lab results, microbiome assays, genetic reports, and NIH-validated PROMIS short-forms — all pulled together on the device, in your possession, before anything is encrypted or transmitted.
DNA reports surface variants that change what every other recommendation should mean for this person — folate cycle, lipid handling, methylation, caffeine clearance, exercise response.
Continuous physiology — HRV, resting heart, sleep stages, recovery, respiration, body temperature, daily activity load. The minute-by-minute reaction to everything else we measure.
CRP, ApoB, fasting glucose, HbA1c, vitamin D, B12, iron, lipid panels. The objective check against the wearable signal — what your physiology is actually doing under the skin.
Gut composition shapes inflammation, mood, nutrient absorption, and how a person responds to any dietary protocol. The map between food, stress, and outcome only makes sense with it.
PM2.5, ozone, pollen, UV, indoor humidity, water quality, ambient noise. The boundary condition every cell in the body is reacting to in real time — and the one most people never see.
Stress, sleep quality, mood, diet logs, supplement stack, work load — captured via NIH-validated PROMIS short-forms. The user's own annotation of why this week looks the way it does — context the data alone won't surface.
Most health apps tell you what was true in 2018. We tell you what was published this week. Recommendations re-evaluate every time a new paper, GWAS hit, or regulatory update lands in one of the corpora below.
~37M biomedical records · NLM open data · live queries
~250M scholarly works · open-access first ranking · full text where licensed
Variant-trait associations · curated genome-wide hits · EBI/NHGRI
Variant-clinical-significance mapping · NCBI · daily updates
US Preventive Services Task Force · grade A/B recommendations
Global clinical guidance · WHO Guidelines · cohort & trial data
UK National Institute for Health and Care Excellence · clinical pathways
~4,500 systematic reviews · evidence weighting · meta-analysis
An agent watches the public science for you — across all eight corpora above. The model only reads public papers. Your encrypted profile never reaches the model. The match between a paper's findings and your profile happens under FHE — cryptographically, not by sending your data anywhere.
Click any stage. The architecture, the encryption state, and the phone respond together. Auto-plays on first scroll.
Keep your Oura. Keep your Apple Watch. Keep your Whoop. We connect to all of them — and add what they can't see. Three axes where the consumer health stack breaks down — and what we do instead.
Every player in the category encrypts at rest, then decrypts the moment they want to look at it. A breach is permanent: you can rotate a password, not a genome.
Wearables know sleep. DNA kits know genes. Labs know markers. No single tool fuses six layers against your local environment in real time.
In the consumer health stack, you are the data and the data is the asset. We invert that — your key, on your hardware, is the only way the insights become readable.
A composite case from the litepaper. Values illustrative, methodology real.
David sleeps seven hours, exercises four times a week, eats well. He also lives where ambient PM2.5 routinely exceeds the WHO guideline by 5×, carries an MTHFR C677T (TT) variant that blunts folate-to-5MTHF conversion (his homocysteine runs high on a "normal" diet), and an APOE ε3/ε4 that makes him cardiovascularly sensitive to the air his commute pipes into his lungs. No single app sees both. VitaCrypt does.
At no point did David's genotype, his HRV time series, or his location-bound PM2.5 exposure exist on a VitaCrypt server in a form anyone could read. The economics of a breach collapse when ciphertext is all that lives on the other side.
A single living profile means the answer changes when the inputs change. Here is David — same biology, same person — across three moments of the same week. Each card is the action VitaCrypt actually puts in front of him.
Depth before breadth. Each phase compounds on the last.
Encrypted ingestion + closed alpha for the first ~50 design partners. DNA, wearable, environment online.
FHE recommendations engine live. Microbiome integration. Invite-only waitlist activated.
Open enrollment + academic-medical-center partnership for IRB-supervised longitudinal validation.
HL7/FHIR connectors, clinician decision support, optional HIPAA BAA path for partner providers.
Apple shipped homomorphic encryption (BFV) on every iOS 18 device in 2024 — the same crypto family we use. Zama's encrypted-genome ancestry benchmark hit 96% accuracy in ~300s on a single CPU (Zama bounty #95, 1000 Genomes test set). And the 23andMe breach finished the job of teaching consumers that "we encrypt at rest" is not a privacy story.
VitaCrypt opens to its first design partners in Q3 2026. Investors and prospective partners — reach out directly.