Privacy Policy

1. Introduction

Libro ("we", "our", or "us") is committed to protecting your privacy and the privacy of your users. This Privacy Policy explains how we collect, use, disclose, and safeguard your information when you use our API, SDKs, and dashboard (collectively, the "Service").

2. Data We Process (And Data We Don't)

Because of Libro's unique Zero-Cost Local Vectorization architecture, the way we handle your users' conversational data is significantly more secure than traditional API-based vector databases:

• No Third-Party AI APIs: We do not send your raw conversation strings to OpenAI, Anthropic, or any third-party embedding provider. All vectorization happens locally within your serverless environment using open-source models.
• Encrypted Storage: Memory vectors and metadata are stored in our secure, SOC2-compliant PostgreSQL infrastructure.
• Data Ownership: You retain full ownership of all data ingested into Libro. We claim no rights to use your data for training generalized AI models.

3. Information We Collect from Developers

When you create a developer account to use Libro, we collect:

• Account Information: Name, email address, and authentication data via Google Auth.
• Payment Information: Processed securely via Stripe (we do not store full credit card details).
• Usage Data: API request volumes, latency metrics, and error logs for debugging and billing purposes.

4. GDPR and Data Deletion

We provide full support for General Data Protection Regulation (GDPR) compliance. If your end-user requests data deletion, you can utilize the ctx.forget(userId) method in our SDK. This performs a hard, cascading delete of all vectors, profiles, and metadata associated with that specific endUserId across our entire infrastructure.

5. Contact Us

If you have any questions or concerns about this Privacy Policy or our data practices, please contact our security team at joshimohanlalit1303@gmail.com.