Data location is vague
Check whether the data processing agreement lists sub-processors outside the EU, and where AI features actually call their models.
EU-hosted CLM with data residency in Germany
Treasure Contracts stores contracts on AWS in Frankfurt, runs AI in EU regions without training on your data and separates every customer in its own database schema.
Two-factor authentication, Microsoft Entra ID SSO, roles and an audit trail with CSV export are part of the product, not a premium add-on project.
ISO 27001 · Hosted in Frankfurt · GDPR · 4-week right of withdrawal
In short
GDPR compliant contract management software needs EU data residency, controlled AI processing, strict access control and traceability. Treasure Contracts stores data on AWS in Frankfurt, processes AI within the platform in EU regions without training on customer data, isolates each tenant in its own database schema and logs activity in an exportable audit trail. Compliance itself also depends on your own processes and agreements.
Key facts
The problem
Employment clauses, contact details, signatures, pricing, health and safety data: a contract repository is one of the most sensitive data stores a company has.
Check whether the data processing agreement lists sub-processors outside the EU, and where AI features actually call their models.
Is the contract text used to improve the model? Which region processes it? Can we choose the provider? Ask each vendor for precise, written answers.
Everyone in the tool sees everything, or permissions have to be maintained by hand. Check whether SSO and a second factor are supported.
When the auditor or works council asks who opened a contract or which AI result was used, the answer is buried in logs nobody can export.
Your outcome
The aim is a tool whose data flows can be explained on one page, and whose controls are visible in the product.
Your team manages, reviews and signs contracts with AI support while data is stored in Frankfurt, AI is processed in EU regions and access is logged in the audit trail.
The controls are product features you can inspect in a demo: tenant isolation, EU model regions, 2FA, SSO, roles and the audit trail.
In 30 minutes we walk through hosting, AI processing and access control on the live product. Open questions from the demo are answered in writing afterwards.
SaaS from Frankfurt, no local installation. Sign-in via Microsoft Entra ID, so existing identities and offboarding processes keep working.
Features
Every control listed here is live in the product today.
The platform and document storage run on AWS in Frankfurt. Documents, extracted fields, versions and signed PDFs are stored in Germany.
AI runs via AWS Bedrock EU, Google Vertex EU or Azure OpenAI, selectable per tenant and work step. Customer data is not used to train models, and every AI call is tracked for tokens and cost.
Each customer has its own database schema. Keys and certificates are stored in encrypted fields.
Two-factor authentication with TOTP or passkeys (WebAuthn), email code as a fallback, and single sign-on via Microsoft Entra ID using OIDC with PKCE.
A role and permission concept controls who can see, edit, review and approve. Approval steps with defined approvers enforce the four-eyes principle (Approval, €190/month, included from Business).
Activity on documents, matters, chats, flows, playbooks and approvals, plus login events, is logged and can be exported as CSV for audits and internal reviews.
How it works
A 30-minute session on the live product: hosting, model regions, tenant isolation, sign-in and audit trail.
Your data protection and IT security questions are answered in writing, including open points from the walkthrough.
Connect Microsoft Entra ID, define roles and choose the AI provider and region that fits your policy.
Two contract types, 10–20 historical contracts, live review with audit trail and acceptance by your team within ten working days.
Comparison
What typical setups mean for data location, AI processing and traceability.
| Criterion | Shared drive and email | Other CLM tools (questions to ask) | Treasure Contracts |
|---|---|---|---|
| Data location | Wherever the files are copied | Where are data and backups stored, and by which sub-processors? | AWS Frankfurt |
| AI processing | Depends on which tools staff use | Which provider and region process AI, and is data used for training? | EU regions, no training, provider selectable |
| Tenant isolation | Not applicable | How are customers separated? | Own database schema per customer |
| Sign-in | Varies | Are SSO and 2FA supported? | 2FA with passkeys, Entra ID SSO |
| Audit trail | File server logs, if configured | What is logged, and can it be exported? | Documents, AI, approvals, logins; CSV export |
General orientation, varies by vendor. Hosting and data handling of other providers vary by product, plan and contract; check the current terms of each vendor.
In detail
Many vendors state that their data centre is in the EU. Ask whether the same applies to every processing step. For contract management, that includes OCR, search indexing, AI review, summaries and chat. Within Treasure Contracts, the platform and storage run on AWS in Frankfurt, and AI is processed in EU regions via AWS Bedrock EU, Google Vertex EU or Azure OpenAI. The model provider can be set per tenant and per work step, so it can follow your internal policy instead of the vendor's default. If you connect ChatGPT or Claude via MCP (AI tool integration, €50/month, included from Professional), that assistant processes the content it retrieves under its own provider's terms.
The most frequent objection from data protection officers is whether contract text ends up improving someone else's model. In Treasure, customer data is not used for training. Every AI call is logged with tokens and cost, which gives you a factual basis for internal reporting. Because AI results in chat are grounded in verbatim quotes from your documents, you can also show reviewers what the model relied on, rather than asking them to trust a black box.
Technical and organisational measures are where GDPR becomes concrete. Treasure separates each customer at database schema level and stores keys and certificates in encrypted fields. Users sign in with a second factor, TOTP or passkeys, or through Microsoft Entra ID single sign-on with OIDC and PKCE. A role and permission concept limits who sees and edits which contracts, and approval steps with defined approvers support a documented four-eyes principle for sensitive contracts.
Accountability under the GDPR means being able to show what happened. Treasure logs activity on documents, matters, chats, flows, playbooks and approvals, together with login events, and exports the audit trail as CSV. Information security is ISO 27001 certified. What software cannot do is make your organisation compliant on its own: your records of processing, data processing agreements, retention rules and internal policies remain your responsibility, and we support that assessment with documentation.
Your first step
You don't have to take our word for it. You see Treasure on your own contracts — first in a free 30-minute session, then in a 10-day Proof Sprint with measurable results.
For GDPR compliant contract management, that means: we walk your data protection officer and IT security through hosting, AI regions and access control, then run two contract types live with full audit trail in the Proof Sprint.
1
30 minutes with the founding team. Bring a real contract and we show Treasure live on your clauses. No slides, 24-hour response guarantee — including a free AI Readiness Assessment.
Pick a time2
€4,900 net, one-off · fully credited towards implementation if you sign within 30 days
Results guarantee: If Treasure misses the acceptance criteria agreed in writing up front, we keep working for up to 20 additional business days at no cost. If that is still not enough, we refund the sprint fee. This requires your timely cooperation; details are in the sprint offer.
Right of withdrawal: 4 weeks unconditional right of withdrawal with a full refund.
We run at most 3 Proof Sprints per month — we confirm the next available start date in the first call.
See pricing & Proof SprintFAQ
Treasure is built for GDPR requirements: data on AWS in Frankfurt, AI in EU regions, no training on customer data, schema-per-tenant isolation, 2FA, SSO, roles and an exportable audit trail, with ISO 27001 certified information security. Whether your use is compliant also depends on your own agreements and processes, which we support with documentation.
AI runs in EU regions via AWS Bedrock EU, Google Vertex EU or Azure OpenAI. You can select the provider per tenant and per work step, for example to stay with an existing Azure commitment. Customer data is not used to train models, and every AI call is recorded with its tokens and cost.
Uncontrolled chatbot use raises data protection questions in contract work: whether there is an audit trail, how data is retained and which rules apply. Within Treasure, AI runs in an EU-hosted, logged environment. If your policy allows ChatGPT or Claude, the Treasure MCP server (AI tool integration, €50/month, included from Professional) lets teams reach Treasure from there with OAuth2 sign-in; the assistant then processes the content it retrieves under its own terms.
Yes. Microsoft Entra ID single sign-on is supported via OIDC with PKCE. Two-factor authentication works with TOTP apps or passkeys (WebAuthn), with an email code as a fallback. Roles and permissions then control who can access which contracts and functions, and login events appear in the audit trail.
Usually one walkthrough plus your questionnaire. We show the controls on the live product in 30 minutes and answer open questions in writing. Because Treasure runs as SaaS from Frankfurt, there is no local installation to assess, and SSO means no separate user administration to review.
Starter €590, Professional €1,790 and Business €3,590 per month, each billed annually and net; Enterprise on request. Approval (€190/month) and Signature (€400/month) are included from Business. The 10-day Proof Sprint costs €4,900 and is fully credited towards implementation if you book within 30 days. A 4-week right of withdrawal with full refund applies.
Read next
In 30 minutes you will know whether Treasure can apply your rules — on a real contract, with the founders.
Free · PDF guide
A practical workbook with ready-to-use prompts for requirements, vendor selection, rollout, adoption and ROI.
Double opt-in, unsubscribe anytime. No sales call, no spam.