Security

Last updated June 2026

UnitCheck is pre-launch. This page describes how the platform is built to protect your operating data. We will share fuller documentation with early-access partners as we onboard.

Tenant isolation

Every customer's data is isolated at the database level with row-level security, enforced in the data layer — not just in application code. One tenant cannot read another tenant's rows.

Human-approved, audited writes

The AI never changes your records on its own. It prepares work with a rationale and the evidence behind it; a person approves before anything is written; and every write is logged and reversible.

Bring your own model

AI inference runs on your own provider account and key, under your governance. You choose the model and hold the spend, and your data is not pooled into a model we control.

Audit trail

Platform actions emit an event log, so there is a record of what happened, when, and on whose approval.

Data in transit

The site and the application are served over HTTPS, and data is encrypted in transit.

What we are still building

We are pre-launch. Formal third-party attestations (such as SOC 2) are on our roadmap, not yet in place — we would rather tell you plainly than imply otherwise.

Contact

Security questions or responsible disclosure? hello@unitcheck.pro.