I have a collection of Egypt-based bank statements and want to turn them into clear, data-driven budgeting insights. The core of the R&D is transaction analysis—specifically extracting each month’s totals and presenting them as an at-a-glance summary that helps users understand where their money goes. Here’s what I need: • Accurate parsing of Arabic/English statement PDFs (OCR or native text). • Reliable classification of transactions into meaningful spending categories. • Automated generation of a concise monthly summary that highlights income, outflow, net balance, and any notable spikes. • Clean, well-commented code (Python with pandas, NumPy, or a comparable stack suits me best) plus a short README showing how to run the pipeline on new statements. Acceptance criteria: the script processes at least three sample statements end-to-end and outputs a CSV or JSON file plus a readable dashboard/report for each month. If you have experience with financial data modelling or existing libraries that speed up statement parsing for Egyptian banks, all the better—I’m happy to iterate quickly on any proof of concept you provide.