Privacy Policy
Privacy Policy for our Tatry Retrieval SDK
This privacy policy explains how our SDK handles data during the retrieval and processing of content for Large Language Models (LLMs).
Data Handling
Our SDK is designed with privacy-first principles. We do not store or retain any content retrieved through the SDK on our servers. The SDK operates as a pass-through service, facilitating retrieval between your specified data sources and your LLM implementation.
Information Processing
The SDK processes:
- Source content metadata
- Attribution information
- Cache configurations
- Query patterns for optimization
- Usage metrics for system performance
We do not:
- Store retrieved content
- Track individual user queries
- Collect personal information
- Monitor LLM outputs
Data Security
The SDK includes built-in security features:
- End-to-end encryption for data in transit
- Secure API authentication
- No persistent storage of retrieved content
- Configurable caching controls
Usage Analytics
We collect anonymous usage metrics solely for:
- System performance monitoring
- Service reliability
- API optimization
- Technical troubleshooting
All metrics are aggregated and contain no identifying information about specific queries or content.
Third-Party Services
When you connect to external data sources through our SDK:
- You are subject to their respective privacy policies
- We do not share information between data sources
- We maintain separate authentication for each source
- You control which sources to connect and access
Updates to Privacy Policy
We will notify users of significant changes to this privacy policy. Continued use of the SDK constitutes acceptance of the updated policy.