Trust & Security
Security you can verify, not take on faith
RigSpec handles revenue, security, and compliance workflows for the businesses that run on it. Here is how we protect that data — grounded in what the platform actually does today.
Tenant isolation
Every record carries an organization key, and every query is scoped to the organization we resolve on the server from your signed-in session — never from a value your browser can set. A database-level backstop is staged underneath it and is not yet switched on; we describe exactly where that stands below.
Encryption & key vault
Data is encrypted in transit and at rest. Sensitive secrets and connector credentials are sealed in a key-managed vault rather than stored in plain application records.
Tamper-evident audit chain
Security-relevant actions are written to a hash-chained audit log. Each entry is linked to the one before it, so any after-the-fact edit or deletion is detectable.
Governed Ai actions
A deterministic policy engine runs before every Ai tool call. The model has no input into it and can never override it — money, security, and data-export actions are never autonomous.
Human-in-the-loop
High-impact actions require explicit human approval and run against a bounded action budget. Money movement and security containment go further than approval: those actuators are not built. Their handlers refuse to run, so there is nothing an approval — or a model — could release.
Consent ledger
Marketing and messaging consent is recorded with its source and timestamp, and honored on every send — opting out suppresses further messages of that type.
Data rights (GDPR / CCPA)
We support data-subject access and deletion. You can request a structured export of your data or its erasure, subject to records we must retain for legal and billing reasons.
Incident & recovery readiness
We maintain incident-response runbooks and a disaster-recovery plan. High-risk Ai actions wait for a person to approve them, the approval screen shows what the action would change before you allow it, and you can stop your Ai assistants at any time from your workspace settings.
Where we are honest about the gaps
A trust page is only useful if it states what is not yet true.
- Database-level tenant isolation is staged, not switched on
- Isolation is enforced in the application layer today: every query is filtered by the organization resolved from your session. We have also written and tested Postgres row-level security policies as a second, independent boundary — but they are not in force yet. The service still connects to the database as the role that owns the tables, and Postgres exempts a table owner from its own policies unless they are explicitly forced. Switching that on needs a least-privilege database role and a change to every query path; it is in progress. Until it ships we will not describe row-level security as an operating control.
- Certifications
- We do not currently advertise a SOC 2, ISO 27001, or HIPAA attestation. We will publish any formal certification here once it is complete — we would rather under-claim than mislead.
- Subprocessors
- Our current list of subprocessors — hosting, database, storage, payments, email and SMS, Ai model and embedding providers, key management, background jobs, and error monitoring, including one provider located outside the United States — is published in full on our Subprocessors page, and summarised in section 5 of our Privacy Policy.
- Service status
- Current operational status for each component is on the status page.
Report a vulnerability
Found a security issue? We want to hear about it. Email security@rigspec.ai or use our contact form. Our machine-readable contact details follow RFC 9116 (security.txt). Please give us a reasonable window to remediate before any public disclosure.
Reviewing RigSpec for your business? Read our Privacy Policy and Terms of Service.
Talk to us about security