Privacy Policy

1. Information We Collect

When you use the Retrivora AI SaaS dashboard, SDK (@retrivora-ai/rag-engine), or LLM Gateway (llm.retrivora.com), we collect the following categories of information:

• Account & Billing Information: Name, email address, project identifiers (projectId), and license subscription status.
• Developer Telemetry Logs: Request status codes, latency metrics, model routing selections, and token counts necessary for quota enforcement and observability.
• Network Identifiers: IP addresses (IPv4/IPv6) used strictly for security rate limiting and fraud prevention.
• Vector Embeddings: Embeddings and metadata generated during document ingestion, stored in tenant-isolated namespaces in Pinecone.

2. How We Process & Protect Your Data

We process information strictly to provide, maintain, and optimize Retrivora AI platform services:

• Zero-Retention Gateway Proxying: Prompt payloads and LLM completions passed through llm.retrivora.com are proxied ephemerally in-memory and are not stored in gateway log databases.
• Encryption in Transit & at Rest: All web requests require TLS 1.3 encryption. Upstream API keys and license signatures are encrypted in backend storage.
• Isolated Vector Spaces: Your document embeddings are partitioned by workspace project IDs (retrivora-<projectId>).

3. Third-Party Service Providers

Retrivora AI integrates with enterprise cloud infrastructure providers to fulfill service requests. Data processed by third parties is bound by strict privacy and confidentiality obligations:

4. Your Privacy Rights (GDPR / CCPA)

Under global data protection laws (including GDPR and CCPA), developers and accounts have the right to:

• Request access to or copies of your account and project data.
• Request permanent deletion of your account, license keys, gateway configs, and vector namespaces.
• Revoke API access keys at any time via the Retrivora Developer Dashboard.