Enhance data processing by adding CSV and Pickle file upload support, improving error handling, and refining the prediction pipeline. Update the `process_records_to_df` function to handle existing DataFrames and ensure required fields are processed correctly.