Security
Secure by design
WrenAI prioritizes robust security measures to safeguard user data.
Notably, users can trust that their database contents are never transmitted to the Language Learning Model (LLM). Instead, only metadata such as schemas, documentation, and queries are utilized in semantic search operations.
WrenAI strictly adheres to a read-only policy for data, ensuring that no sensitive information is stored or transmitted unnecessarily. In instances where data preview is required, queries are submitted on-demand, and no preview data is retained within WrenAI's database or storage systems, reinforcing user privacy and data integrity.