abdullahiabuzaid70-pixel/revai

RevAI - AI Revenue & Fraud Detection for African Companies # RevAI - AI Revenue & Fraud Detection for African Companies > Upload your financial data. We find your leaked money. You pay us 10% of what we recover. ## What It Does RevAI scans company financial data (AP, Payroll, Tax) and finds money quietly leaking out through 5 automated detections: 1. **Duplicate Payments** - Same invoice paid twice, or near-duplicate amounts to same vendor within 30 days 2. **Unremitted VAT/WHT** - VAT/Withholding Tax deducted but never remitted to FIRS 3. **Overstated Expenses** - Abnormally high spending, round-number fraud patterns, statistical outliers 4. **Ghost Vendors/Employees** - Payments to entities not in master lists, shared bank accounts, missing PAYE/pension 5. **Late Tax Filing Risk** - Missed filing deadlines with estimated penalty exposure ## Quick Start ```bash # Install dependencies pip install -r requirements.txt # Add your OpenAI API key cp .env.example .env # Edit .env and add your key # Run the app streamlit run app.py ``` Then click "Use Sample Data (Demo)" in the sidebar to instantly load realistic Nigerian company data with known fraud patterns. ## Tech Stack - **Frontend:** Streamlit (Python web app framework) - **Data Processing:** Pandas, NumPy - **Vendor Matching:** FuzzyWuzzy (fuzzy string matching) - **AI Summary:** OpenAI GPT-4o-mini (optional - falls back to template if no API key) - **PDF Generation:** fpdf2 - **Tax Rules:** Hardcoded Nigerian tax rates (VAT 7.5%, WHT by category, PAYE brackets, penalties) ## File Structure ``` revai/ app.py # Main Streamlit application requirements.txt # Python dependencies .env.example # Environment variable template detectors/ duplicate_payments.py # Detection 1: Duplicate payments vat_wht_remittance.py # Detection 2: VAT/WHT issues overstated_expenses.py # Detection 3: Overstated expenses ghost_vendors_employees.py # Detection 4: Ghost vendors/employees tax_filing_risk.py …