Tenant isolation by default
Every row carries a tenant_id and Postgres Row-Level Security enforces who can see what. There is no application-only filter that an SQL injection could bypass.
Security & trust
Built for HIPAA. Designed for auditors.
MedIQ was architected as a multi-tenant healthcare platform from line one. Every layer — database, application, AI — enforces tenant isolation and the principle of least privilege.
Encryption at rest & in transit
TLS 1.2+ everywhere. Database storage encrypted at rest. Any optional tenant secrets are encrypted with AES-256-GCM using per-record IVs.
Role-based access
Super admin, tenant admin, biller, provider, front desk, and auditor roles. Sensitive surfaces (secrets, billing, role changes) gated server-side.
Audit log of every change
Status transitions, role grants, secret rotations, scrubber decisions — all written to an append-only audit log keyed by tenant.
No PHI in AI prompts
Prompts carry billing codes, amounts and service dates — never names or member IDs. Note text is scrubbed first. Logs record the decision and its data class, not identifiers.
BAA + SOC 2 (in progress)
Signed Business Associate Agreement available before any PHI flows. SOC 2 Type II audit in progress; report available under NDA.
Backups & DR
Point-in-time recovery up to 7 days on Growth, 30 days on Enterprise. Tested restore drills quarterly.
Infrastructure
Hosted on enterprise cloud providers with SOC 2 / HIPAA-eligible services. Dedicated VPC option on Enterprise.
Responsible disclosure
We publish a security.txt and operate a coordinated-disclosure program. Email security@cybergenai.com.
Engineering posture
Defense in depth, applied to billing data.
Every change is code-reviewed, every dependency is automatically scanned, and every deploy is reproducible. Migrations are versioned and applied with idempotent SQL.
- Mandatory MFA for all staff
- SSO (SAML/OIDC) on Enterprise
- Secret scanning on every commit
- SAST + dependency scanning in CI
- Production access reviewed quarterly
- Vendor reviews tracked in writing
- Penetration test annually
- Incident response runbook + on-call
AI security
How AI works in MedIQ.
Questions we hear most often about model choice, data handling, and how tenant secrets stay secret.
No. MedIQ centrally vets, approves, and assigns models per AI task. Tenants never select models or hold third-party AI keys — our operations team monitors performance and compliance and adjusts assignments for you.
Need a vendor security review?
We can share our security overview, sub-processor list, and a draft BAA under NDA. Most reviews close in under a week.