Document review
When you submit a PDF, DOCX, or TXT file, the bytes are read in request memory. The application does not write the source file or the report to a database or filesystem. Extracted text and allowlisted structured intake fields are sent to the OpenAI API with response storage disabled. PrenupsAI extracts the file in server memory and does not save it or your report. Extracted agreement text and the structured review context are sent to the OpenAI API with response storage disabled. API data is not used for model training by default, but content may be retained for up to 30 days for abuse monitoring.
Intake fields are structured flags and dates only. Do not enter names, addresses, account numbers, balances, or a freeform narrative. Request logs may contain a random request ID and a normalized error message. They must not contain document text, filenames, or report bodies.