HJ-tech205/bee_money-wise

Personal finance analytics and investment guidance for South African students. # BEE Money-wise A personal finance analytics platform focused on South African students and young earners. It ingests bank statement PDFs from multiple South African banks, breaks down spending and income, teaches budgeting, compares debt repayment strategies, and generates tax-aware investment guidance. Repo: github.com ## What it does - **Ingests bank statements** from multiple SA banks (FNB, Standard Bank, Capitec to start) via PDF parsing — no manual CSV exports needed. - **Analytics** — spending by category, month, and merchant; income vs expense trends; recurring expense detection (subscriptions, regular bills). - **Budgeting guidance** — framework-based recommendations (e.g. 50/30/20) applied to your actual categorized spending, not generic advice. - **Debt payoff optimizer** — compares avalanche vs snowball repayment strategies with real rand and time savings on your actual debts. - **Tax-aware investing** — given an amount and timeline, applies current SARS rules (TFSA limits, interest exemption, dividends tax, CGT) and runs a portfolio optimizer (mean-variance) to suggest an allocation. - **Credit score education** — an interactive estimator that teaches the real factors behind SA credit scoring (there's no public API for pulling a real score from a third-party app, so this is deliberately educational rather than a live data pull). ## Setup ```bash python3 -m venv venv source venv/bin/activate # Windows: venv\Scripts\Activate.ps1 pip install -r requirements.txt ``` If `prophet` or `camelot-py` fail to install (both have system-level build dependencies), comment them out of `requirements.txt` and get the core pipeline running first — pandas, pdfplumber, scikit-learn, and Streamlit cover the MVP without them. ## Project structure ``` src/ ingestion/ PDF parsers, one per supported bank analytics/ cleaning, categorization, recurring-expense detection budgeting/ budgeting framework logic debt/ …