The community member is looking for techniques to use for unstructured PDF data in their RAG chatbot. Another community member suggests trying LlamaParse, which is described as a state-of-the-art PDF parser. However, the original poster is concerned about data privacy, as the PDF data is sensitive. The community members discuss that LlamaParse only keeps the data for 48 hours in the cloud, and it is removed after that time period, addressing the data privacy concern.
No data is kept only for 48 hrs in the cloud for reason such as if you try to call the API again it would not take time like it did for the first time and return it quickly. After 48 hrs data is removed from the cloud.
