Python Programmer

We are a specialized insurance agency looking for an experienced Python developer to build a backend automation script that processes our monthly commission statements.
Our carriers send us statements in wildly different formats (PDFs, CSVs, and ZIP folders). We need a script that takes a batch upload of these files, uses the Google Gemini API to intelligently extract the data, and standardizes it into a strict CSV format required by our CRM.
The Good News: You won't have to start from scratch. We have already written the core API interface code and the complex prompt engineering required to parse the unpredictable PDFs. We will provide you with a comprehensive Project Brief containing this code, the exact data formatting rules, and a ZIP file of dummy statements for you to test against.
Your Responsibilities:
File Handling: Write the logic to accept a batch of files (or extract a .zip) and feed them to the API loop one by one.
User Interface: Wrap the script in a simple, lightweight GUI (like Streamlit or Tkinter) so our team can drag-and-drop files to process them without using a command line.
API Integration: Implement our provided Gemini API starter code, forcing the model to return structured JSON. Handle API rate limits gracefully (e.g., using time.sleep()).
Data Formatting: Write the Python logic to enforce strict formatting rules (e.g., ensuring all extracted dates are formatted strictly as MM/DD/YY, calculating a "Total" row, and flagging missing data).
Dynamic Export: Configure the script to output separate, dynamically named CSV files for each statement processed (e.g., [Company Name] [Month] [Year].csv).
Security: Ensure the script automatically deletes processed files from Google's servers immediately after generation to protect client data.
Requirements:
Strong proficiency in Python.
Experience working with the Google Generative AI SDK (google-generativeai).
High attention to detail (our CRM requires exact column headers and date formats).
Project Scope:
This is a well-documented, straightforward project. We estimate it will take a competent developer 4 to 8 hours to complete and thoroughly test against our dummy data suite.
This will act as a paid test project. If the code is resilient and the instructions are followed perfectly, we have ongoing freelance development work available.