Privacy and data handling
Documents are processed for the request. The Document Intelligence pipeline does not persist raw uploads by default, and extracted OCR text is not stored in API usage telemetry. Raw provider responses are not exposed through the API.
Operational logs and telemetry use safe metadata such as request ID, status, page count, bytes, latency, and provider invocation count where applicable. Retain request_id for troubleshooting without placing document content in support tickets.
For OERENA’s broader privacy information, read the Privacy Policy.