Product architecture
How Vitalis works under the hood
A look at the core flows — intake, AI generation, practitioner review, fulfillment, and ongoing telemetry — and the rules that connect them.
Core flow
Data, decisions, and actions across the member journey.
Intake
Member completes a guided assessment
AI Engine
Signals → personalized protocol draft
Review
Practitioner approves or adjusts
Activate
Protocol + supplements assigned
Track
Daily adherence + outcomes loop
Protocol Rules Engine
A deterministic + ML hybrid that proposes protocols, then defers to a clinician.
Inputs
- Assessment scores (1–100)
- Wearable telemetry (HRV, sleep)
- Lab markers (when uploaded)
- Preferences & contraindications
Decision logic
IF sleep_score < 70 AND magnesium_low THEN suggest "Calm Magnesium Glycinate" AND assign "Sleep Reset" protocol PRIORITY = high IF stress_score > 60 AND cortisol_pm_high THEN suggest adaptogen stack AND schedule weekly check-in
Outputs
- Personalized protocol draft
- Supplement recommendations
- Reminder cadence
- Practitioner review task
System diagram
┌─────────────┐ ┌──────────────┐ ┌────────────────┐
│ Member App │──────▶ │ Assessment │──────▶ │ Rules Engine │
└─────────────┘ │ Service │ │ + AI Model │
└──────────────┘ └────────┬───────┘
│
▼
┌─────────────┐ ┌──────────────┐ ┌────────────────┐
│ Practitioner│◀────── │ Review Queue │ ◀───── │ Protocol Draft │
│ Console │ └──────────────┘ └────────────────┘
└──────┬──────┘ │
│ approve │
▼ ▼
┌─────────────┐ ┌────────────────┐
│ Protocol │────────────────────────────────▶│ Notifications │
│ Activation │ │ + Commerce │
└─────────────┘ └────────────────┘
│
▼
┌─────────────┐ ┌──────────────┐ ┌────────────────┐
│ Member runs │──────▶│ Telemetry & │──────▶ │ Outcomes loop │
│ daily steps │ │ Check-ins │ │ feeds engine │
└─────────────┘ └──────────────┘ └────────────────┘Core entities
Member
Profile, preferences, history
Assessment
Versioned intake responses + scores
Protocol
Goals, steps, cadence, adherence
Order
Cart, shipment, subscription state
Clinician
Roster, credentials, schedule
Event
Audit trail of every change
Consent
Granular sharing preferences
Model run
Inputs/outputs traceable to a version
Want to explore the workspace?
Sign in to step into a member, practitioner, or admin workspace.