Rilev Docs
A clear technical overview of the Rilev platform: the app surfaces, API families, anonymity model, zero-knowledge architecture, and the boundaries we preserve when building integrations.
Overview
Rilev is a psychological assessment platform with three product surfaces: an anonymous individual app, a professional workflow layer, and an enterprise aggregation platform. The technology underneath is shared, but the trust model is intentionally different for each audience.
The individual platform is designed around a strict separation between account identity and assessment data. Professional and enterprise systems add operational layers on top of that model without weakening the underlying anonymity guarantees.
Docs Library
These focused pages make the platform easier for people and crawlers to understand. Each page has its own canonical URL, topic-specific metadata, and structured breadcrumb context.
Platform Guide
How the individual app, professional platform, enterprise layer, and integration surface fit together.
API Guide
A crawler-readable explanation of the main API families, public versus internal surfaces, and integration boundaries.
Zero-Knowledge Guide
The technical privacy model behind capability-based access, one-way ownership proof, and plane separation.
Anonymity Guide
A concise map of account anonymity, storage separation, and what Rilev intentionally does not collect.
Security Model
How Rilev layers session verification, request-origin protection, body limits, abuse controls, ownership checks, and provider verification.
Data Lifecycle
What data is created, where it lives, how it moves, what is intentionally minimized, and where deletion boundaries sit.
Integrations Guide
How GPT Actions, MCP-style tools, live context, authorization grants, and provider callbacks fit into the Rilev trust model.
Privacy Impact Assessment
Formal assessment of data processing proportionality, risks, safeguards, and individual rights — required under Quebec Law 25 and best practice under PIPEDA.
Data Processing Agreement
Standard DPA for professional-tier users outlining data processing obligations, breach notification, subprocessor disclosure, and cross-border transfers.
Product Map
Individual App
Anonymous assessment, personal dashboard, progress tracking, and report generation built around access-key accounts instead of identity-first signup.
Professional Platform
Referral codes, client report workflows, Stripe-gated report generation, sharing controls, and operational analytics for professionals.
Enterprise Platform
Assessment waves, aggregated wellness reporting, departments and locations, access control, and privacy thresholds that suppress small groups.
AI & Integration Layer
GPT Actions, MCP-style workflows, live-context grants, OAuth code exchange, and report-generation services with hashed token storage.
API Surface
Rilev's API surface is grouped by platform responsibility. Some surfaces are public product integrations; others are internal application handlers. This docs page explains the domains and design intent rather than publishing every private implementation detail.
Session lifecycleAnonymous access restorationAssessment sessionsDerived outcomesProtected decode flowsDaily check-insTrend readsHealth aggregatesReferral workflowsReport jobsBilling portalsAssessment accessAggregate reportingExportsAI actionsLive contextMachine integrationsPrivacy Architecture
Capability-based data access
Sensitive records are addressed through random capability-style access, not ordinary identity identifiers.
One-way ownership proof
The identity plane stores only a one-way proof. The raw capability is not persisted as a readable account-to-data link.
Separated planes
Identity-plane records and data-plane records are split so normal database reads do not reveal an ownership graph.
Defense-in-depth APIs
Sensitive handlers combine authenticated sessions, request-origin checks, body limits, abuse controls, and ownership verification.
Data Flow
- 1The client authenticates or restores access through the anonymous account flow.
- 2The app resolves the user-held capability locally and presents a one-way proof when calling protected data-plane surfaces.
- 3The server verifies the active session and proves ownership without storing a readable identity-to-data link.
- 4Data-plane records are read or written under anonymous storage locations rather than identity-keyed locations.
- 5Identity-plane metadata remains limited to account and entitlement state, not clinical payloads.
Integrations
GPT Actions
OAuth authorization-code flow, one-time codes, hashed tokens, and professional action scopes.
Live Context
Anonymous context grants that remain session-bound and ownership-checked.
Provider callbacks
Payment and subscription callbacks are isolated from browser-request assumptions and guarded by provider verification.