Skip to main content
Security

Security controlsand procurement evidence.

Tenant isolation, a tamper-evident audit chain, MFA, encrypted backups, AI that never trains on your data, a SOC 2 roadmap, and named sub-processors. We publish how each one works so you can verify it — not just trust it.

For procurement we send current control evidence and the date of the latest review. Hosting and AI service regions are listed in Privacy; regional deployments are agreed separately. Our SOC 2 position is set out below.

Access and identity

Tenant isolation

Postgres RLS with FORCE ROW LEVEL SECURITY on every tenant table.

Every table that holds tenant data is scoped by RLS policies enforced inside Postgres — not at the application layer. We use FORCE ROW LEVEL SECURITY so even the table owner cannot accidentally bypass the policy. Files in object storage sit under your organisation’s own path, and storage policies check membership on every read, with signed-URL access.

MFA

TOTP via Supabase Auth, with step-up for report signing.

Time-based one-time-password authentication is available to every user. Signing and countersigning a report requires an AAL2 step-up authentication check. We do not support SMS-based MFA.

Data handling

Audit chain

SHA-256 hash chain via DB trigger. Mathematically tamper-evident.

Every signable artefact — log, lab result, certificate, report, signature — is hashed with SHA-256 and chained to the previous record by a database trigger that you cannot opt out of. Tampering with any historical record breaks the chain. The chain is re-verifiable from a CSV export, so you do not have to trust our hosting to trust the audit.

AI privacy

AWS Bedrock. Never trained on your data.

AI inference runs on AWS Bedrock in the service region described in our Privacy Policy. The Bedrock foundation models used for reasoning, photo classification, and fast-tier checks are not trained on customer data. Every user-facing AI path runs on AWS Bedrock.

Data ownership

Your data is yours. Full export, anytime, no lock-in.

You own every borehole, lab result, model, and report you put into TerraBrains. Export supported workspace data and files from the application, including AGS 4.x, DIGGS 2.x, CSV, report PDFs, and audit-chain hashes. We never sell your data, never share it for advertising, and never use proprietary borehole models or reports to train AI. Our retention and deletion practices are described in the Privacy Policy.

Encryption

Encrypted in transit and at rest.

All traffic is TLS 1.2+ end to end. Data at rest — Postgres, object storage, and backups — is encrypted with AES-256 by the managed cloud layer. Production secrets and signing keys are managed through secret stores. Request current review evidence during procurement. Authenticated file access uses time-limited signed URLs where implemented. These links remain usable until expiry; treat them as confidential.

Backups & recovery

Daily automated backups.

The database is backed up automatically every day by our database provider, and the last seven daily backups are kept, so we can restore to any of those days. Object storage is redundantly stored across availability zones. The audit chain means a restore is independently verifiable — you can prove the recovered data is the data you signed.

Data residency

Data-location terms are stated before procurement.

The shared service currently uses the primary hosting and AI-inference regions described in our Privacy Policy. A regional or single-tenant deployment may be scoped for Enterprise customers by arrangement; it is not a self-service residency selector.

Assurance and procurement

Compliance

Controls mapped to SOC 2; report in progress.

Our controls — access management, encryption, change management, logging, and vendor review — are mapped to the SOC 2 Trust Services Criteria. A Type II report is on the roadmap; we are not certified. A draft DPA can be requested for procurement review and becomes binding only when agreed by the parties.

Sub-processors

Providers and processing purposes — published.

We use a small, deliberate set of sub-processors: AWS (App Runner + CloudFront for web delivery, Bedrock AI, and storage), Supabase (managed Postgres + auth), Dodo Payments (merchant-of-record billing and tax — it handles worldwide payments, currency and tax), and Resend (transactional email), and PostHog (US cloud — product analytics and session replay, only after a visitor accepts cookies, with form inputs and project content masked). Each is a reputable provider with its own published security posture. We notify customers before adding a new sub-processor that touches tenant data.

On-premise

Single-tenant and on-prem deployments are scoped separately.

Firms with data-localisation or isolated-environment requirements can discuss a single-tenant or on-prem deployment with us. Availability, architecture, support boundaries, and commercial terms are agreed in a separate written scope; these deployments are not included in the self-service product.

Dependencies

Dependency management and review.

Dependency versions are pinned and upgrades are reviewed. Current dependency-review and scanning evidence is available during procurement.

Reporting a vulnerability

Reporting

Found something? Email security@terrabrains.ai.

We review good-faith security reports sent to this address. Please include enough detail for us to reproduce the issue and avoid accessing or modifying other customers’ data. Our acknowledgement target is one business day; this is not a contractual SLA or a remediation deadline.

Responsible disclosure

We aim to acknowledge within one business day.

Email security@terrabrains.ai with a description of the issue, the steps to reproduce it, and the impact you observed. This acknowledgement target is not a contractual SLA. Investigation and remediation time depend on severity and scope; we will discuss next steps with you. Tell us if you would like public credit after resolution.

Judge it on your own project.

Start free and see how records, signatures and access work in practice. Security questions go to security@terrabrains.ai.