Plans, Billing, and Data Security — FAQ

1 views

Frequently Asked Questions

Plans & Pricing

What plans are available?

Plan Price AI Queries Documents Team Size GPU Access
Sandbox Free 100/month 100 1 user No
Professional €45/month 1,000/month Unlimited 3 users Yes
Corporate €349/month 10,000/month Unlimited 50 users Dedicated
Sovereign From €1,500/month Unlimited Unlimited Unlimited Dedicated

Annual billing gives a 20% discount on all paid plans.

Is there a free trial?

Yes. All paid plans include a 7-day free trial with no credit card required. You get full access to the plan's features during the trial period.

What happens if I hit my query limit?

Your account is paused — you won't be charged overage fees automatically. You can either wait for the next billing cycle or upgrade your plan. Usage-based add-ons are available as an opt-in for users who want to pay per extra query (from €0.05/query).

Can I change plans?

Yes. Upgrades take effect immediately and are prorated. Downgrades apply at the next billing cycle.

Data Security & Privacy

Where is my data stored?

All data is stored on bare-metal servers in Germany and Finland (Hetzner data centers). No data leaves the EU/EEA. We do not use AWS, Azure, Google Cloud, or any non-European cloud provider for data storage.

Do you use OpenAI, Google, or other third-party AI APIs?

No. We run our own AI models on our own GPU hardware. Your documents and queries are never sent to OpenAI, Google, Anthropic, or any third-party AI provider. The entire inference pipeline runs on European infrastructure we own and operate.

Is CaveauAI GDPR compliant?

Yes. CaveauAI is designed with GDPR compliance from the ground up:

  • Data minimization — We only process what's needed for the RAG pipeline
  • Tenant isolation — Your data is completely separated from other clients
  • Right to deletion — Delete any document and its vectors are removed from all systems
  • Data Processing Agreement (DPA) — Available for all paid plans
  • No data sharing — Your private documents are never used for model training or shared with other tenants

What about tenant isolation?

Every client's data is strictly isolated:

  • Private corpus: Stored in a dedicated partition filtered by your client ID
  • Vector store: Qdrant points are tagged with your client ID and filtered at query time
  • File storage: Documents stored in client-specific directories
  • Corporate+ plans: Optional dedicated database instance

Can I export my data?

Yes. You can download your documents via the API (GET /api/v2/documents.php) or request a full data export through support.

AI Models & Infrastructure

What AI models does CaveauAI use?

Model Parameters Purpose
bnl-legal 7 billion Fine-tuned for Norwegian and European law. Fast, specialized responses.
Qwen 2.5 72B 72 billion Enterprise-grade reasoning. Handles complex multi-domain queries.
nomic-embed-text — Embedding model for document vectors (768-dimensional).

Model availability depends on your plan. Professional and above get GPU-accelerated inference.

What is the knowledge distillation pipeline?

We use a multi-step process to create domain-specific AI models:

  1. Teacher models (NorwAI-24B and Qwen-72B) generate high-quality Q&A pairs from legal documents
  2. These Q&A pairs are curated and quality-scored
  3. A smaller model (Qwen 7B) is fine-tuned using QLoRA on this curated data
  4. The result is a fast, specialized model (bnl-legal) that punches above its weight on domain-specific questions

What is RAG (Retrieval-Augmented Generation)?

RAG is the technique that prevents AI hallucination. Instead of relying solely on the model's training data, CaveauAI:

  1. Searches your document corpus for relevant passages
  2. Retrieves the most relevant text chunks
  3. Augments the AI model's prompt with these real documents
  4. Generates an answer that cites specific sources

This means every answer can be verified against the original documents.

Account & Access

How do team roles work?

Role Permissions
Owner Full access including billing, team management, API keys
Admin Manage documents, scrapers, team members (no billing)
Editor Upload documents, use chat and tools
Viewer Read-only access to chat and documents

Can I use the same login for billing and the AI platform?

Yes. Your billing portal account at billing.bluenotelogic.com is linked to your CaveauAI account at ai.bluenotelogic.com via Single Sign-On.

How do I get API access?

API access is available on Starter plans and above. Generate your API key from the AI platform dashboard. See the API Documentation article for full details.

Support

How do I get help?

What are the support response times?

Plan Response Time
Sandbox Best effort (community)
Professional Within 48 hours
Corporate Within 24 hours
Sovereign SLA-defined (typically 4 hours)
Go back Similar articles