OpenMined/PySyft ? reverse-engineered prompt

Build me a Python tool for privacy preserving data science where a data scientist can ask to run code on someone else’s private data, but the private files never leave the owner’s side. I want a simple flow where one person can sign in as the data owner, another as the data scientist, share a dataset, and then submit a Python script that runs only after approval. The data scientist should be able to test against mock data first, then send the real job, and get back only the approved output. It should work through cloud storage like Google Drive, be easy to use from local Python or Jupyter, and still keep working if one side is offline and sync later. Please include the basic login, peer approval, dataset sharing, job submission, job approval, and result download flow, with clear examples and a clean README that shows how to try it end to end. If you need current setup details, look up the latest docs online.