How ProcessTwin protects your data

A summary of the protections around your workspace. The full engineering detail lives on our security page.

Isolation

Your workspace is isolated at two independent layers: every query is scoped to your tenant in the application, and PostgreSQL row-level security enforces the same boundary underneath. Automated cross-tenant probes run in our CI on every change — the test suite literally tries to steal data between test tenants and must fail.

Encryption

• In transit: TLS everywhere.
• At rest: disk-level encryption, plus application-level AES-256-GCM for connector credentials.
• Per-tenant keys: credential encryption keys are derived per workspace, so one workspace's ciphertext is useless against another.

Audit

Every privileged action — by a human, an agent, or our own support staff — writes to an append-only, hash-chained audit log. If anyone ever altered history, the chain would break visibly. Support access to your workspace (impersonation) requires a reason, is time-limited, and appears in your audit log too.

Where your data lives

ProcessTwin is a fully managed service running on hardened servers in Germany. We operate under UK GDPR and the Data Protection Act 2018; UK→EEA hosting is covered by the UK's adequacy regulations, and our DPA & subprocessor register lists every third party we use with its transfer mechanism. Enterprise plans can run on a dedicated single-tenant instance for stricter isolation requirements.

What we can't do

We cannot read your connector credentials in plaintext, cannot use your data to train models, and cannot access your workspace without leaving an audit trail you can see.