Clear controls for your laboratory’s data
Every claim on this page maps to something running in the platform today. Not a roadmap item, not a certification we're "working toward." If it isn't built, we say so plainly instead of implying it.
What's really protecting your company's data
Per-Company Data Isolation
Every domain model in Vitrio (inventory, tests, purchase requests, suppliers) carries a required link to exactly one lab. That link is structural, not a filter someone forgot to apply: there's no code path where one company's data can be queried under another company's account.
Ten Real Roles, Least-Privilege by Default
System Admin, Lab Admin, Manager, Technician, Procurement Officer, Front Desk, Viewer, Supplier Portal User, Marketplace Supplier and Referral Partner are enforced server-side on every view, not just hidden in the UI. A Viewer account can't reach an edit endpoint by guessing the URL, and a Front Desk account is structurally confined to its one page regardless of what URL it's given.
A Full, Queryable Audit Trail
Logins, logouts, exports, permission denials, and every create/update/delete are written to an audit log with the acting user, IP address, user agent, and timestamp, reviewable by a Lab Admin, not buried in a server log file.
Brute-Force Login Protection
Both the source IP and the target account are automatically locked out after repeated failed logins, built directly on the audit log. No third-party service to configure, and nothing that can silently fail to fire.
Encrypted Integration Tokens
OAuth tokens for connected platforms like Microsoft 365 and Google Workspace are Fernet-encrypted at rest with a dedicated key, never stored or logged in plaintext.
What happens if a password leaks
This is the real sequence, built on the mechanisms above.
1. Repeated failed logins are detected
Every login attempt is checked against the audit log in real time. There's no delay before the pattern is visible.
2. The account and the source IP are locked out
Automatic, no admin action required: 5 failed attempts from one IP within 10 minutes, or 10 against one account, and further attempts are blocked.
3. The attempt is logged, not silently dropped
Each failed login writes an audit record with the IP address and user agent. A Lab Admin can see exactly what was tried and from where.
4. Tenant isolation holds even if a login succeeds
A compromised account still only ever sees the data belonging to its own lab. Isolation is structural, not a permission that can be bypassed from inside a session.
5. A Lab Admin can review the whole picture
The audit trail isn't limited to this one incident. Every login, export, and permission-denied event across the account is queryable after the fact.
Speak with our team
Security questions are answered by people who know the platform, not a support queue. If something on this page needs more detail for your own review, ask.