Trust & security

Intelligence you can defend.
Data you still own.

CrawlQ Studio is built for teams whose content has to survive Legal, Compliance, and a Board-room question. Here is how we design for that, end to end. Studio is built on GraQle — the architecture layer independently named in a public EU AI Act compliance framework by senior practitioners including Peter Borner (Open Privacy Standards Foundation) and Andrii Matiash (VERITAS Framework).

Where your data lives

European Union by default.

Studio runs on European AWS infrastructure — not a pass-through, not a configurable add-on, the primary deployment.

Data residency

European AWS infrastructure

Your workspace, your documents, your brand map, and every AI response all run from our EU region. GDPR is the starting line, not a bolt-on.

GDPR-native
Data processing agreement available on request. Subject-access requests handled directly by support.
EU AI Act ready
Designed for the governance obligations landing in the EU AI Act, including record-keeping, transparency, and human oversight requirements. (Not legal advice.)
No transatlantic transfers by default
EU customer data stays inside the EU region. Any exception requires explicit customer consent and is documented.
Encryption in transit and at rest
TLS 1.2+ in transit. AWS-managed encryption at rest for storage and database tiers.

How accounts stay yours

Auth built for teams that get audited.

Enterprise-grade identity on day one, not a paywalled add-on.

Identity

Amazon Cognito + TOTP MFA

Accounts are backed by Amazon Cognito. Time-based one-time-password (TOTP) multi-factor authentication is available from day one — compatible with any authenticator app.

Cognito-backed sessions
Industry-standard session tokens. No custom auth rolled in-house.
TOTP MFA available
Turn it on from your profile. Recovery codes are generated and shown once for offline storage.
No forced password resets on migration
Legacy CrawlQ accounts move to Studio without a reset dance. Same email, same credentials.
Least-privilege admin surfaces
Team roles scoped to the workspace. No blanket admin tokens in customer-facing code paths.

Your brand data

Still yours. We do not train on it.

Your documents, your brand map, and the content you generate stay yours. We don't use them to improve our AI. We don't let any model provider use them either.

Data sovereignty

No training on customer data

Customer content does not enter any training pipeline — not ours, not the model provider's. Inference is the full extent of what any model sees.

Zero training on brand data
Your uploads, your conversations with Athena, and your generated outputs do not feed model training. Ever.
Isolated workspaces
Your brand data lives inside your workspace. It is never mixed with any other customer's data.
Deletion is your explicit choice
When you delete content, it is taken out of active storage. Backups age out on a documented retention window.
Export without ransom
Your documents, your outputs, and everything Studio has learned about your brand can be exported at any time. You are never locked in.

How AI stays accountable

Human-in-the-loop, by design.

Every AI output is ranked, evidenced, and reviewable before it ships. The human decides; the platform surfaces the argument.

AI governance

Evidence on every output

Studio is built around the BRAND score and citation-backed reasoning, so every output carries an explanation your Legal team can read.

BRAND score on every output
Five dimensions — fidelity, reasoning, audience alignment, novelty, deliverability. Nothing ships without a score.
Citation-backed research
Athena's answers surface the knowledge-graph entities they drew from, so claims can be traced to their source.
Reversible by default
Drafts are drafts. Approval is explicit. Nothing auto-publishes, nothing auto-sends.
Audit trail
Every generation, every review action, and every Canvas run is logged inside the workspace for your records.

Engineering you can audit

Trust claims, backed by public engineering.

CrawlQ is built on an open-core engineering substrate whose governance machinery is public. You do not have to take our word for the controls below — your engineering or DPO team can inspect every one of them at the file level, today.

EUR-Lex Drift-Guard

Our engineering substrate’s CI reads the regulation text weekly — and flags drift.

A public, scheduled CI workflow re-hashes every EUR-Lex page the substrate’s compliance documentation cites, every week. If the regulator changes a page, an issue opens automatically. When your auditor asks “how do you stay current with the EU AI Act?”, this is an answer in code, not a policy promise.

Rekor Audit Anchoring

Cryptographic audit trails anchored to public infrastructure — they remain verifiable even if we disappear.

Governed decisions in the substrate get a cryptographic receipt — canonical JSON, Merkle tree, ed25519 signature — anchored to the public Sigstore Rekor transparency log. An auditor can verify the records independently, without access to our infrastructure. Vendor-disappearance is priced into the audit design.

Compliance mode is a cryptographic latch
EU-AI-Act mode is an ed25519-signed, hash-chained, irreversible switch. Once enabled it cannot be silently disabled — a tamper attempt fails closed.
13 public EU AI Act article-mapping documents
Article-by-article mapping (Articles 4, 12, 13, 14, 15, 25, 43, 50 and supporting schemas), written to be quoted in your own compliance file. Read them.
Non-claims discipline, enforced in code
A release-blocking invariant test scans every governance record and refuses to ship if a “compliant” or “certified” field appears. The vocabulary discipline your legal team wants is a test suite, not a style guide.

The substrate itself is open-core engineering you can audit: 150k+ lines of Python, 440 public test files (~7,700 test functions), public CI on Python 3.10–3.12, 140+ PyPI releases, and a reproducible MultiGov-30 benchmark harness — verify yourself. Full engineering disclosure → Meet GraQle, the SDK itself →

Three European patent applications pending on the underlying methods (patent-pending). Engineering controls described here are properties of the public substrate; not legal advice.

Legal + procurement

DPA, security reviews, procurement questionnaires.

Send us your questionnaire. We answer directly, in writing, usually within two business days.

Request the Data Processing Agreement

Mention your company name and the products you plan to use. Our DPA template covers the standard GDPR Article 28 posture.

Request the DPA