Privacy Policy
This policy covers the public Pipeline Claims ChatGPT companion.
Data you may provide
The app accepts only a PDF that you explicitly attest is fictional or fully deidentified. You must not provide PHI, real medical records, government identifiers, credentials, or information that identifies a real person or claim.
How data is used
The service downloads the temporary file supplied by ChatGPT, validates it, and extracts bounded native text to answer your request. Document text is treated as untrusted source material. The service does not use the Pipeline OpenAI API key or send the document to another model provider.
Storage and retention
The service does not create user accounts, save uploaded PDFs, or persist extracted document text. Files and excerpts are held only in process memory for the request. The hosting provider may process standard operational metadata such as IP address, timestamps, user agent, and request status for security and reliability.
Sharing
The service does not sell document data. Data is processed by the hosting infrastructure needed to operate the service and by OpenAI/ChatGPT under the terms that apply to your ChatGPT use.
Your choices
Do not use the upload tool if the document cannot be safely fictionalized or fully deidentified. You may use the built-in fictional demonstration without uploading a file.
Contact
Privacy questions: nickvan23@gmail.com.