tseon-designs/ethiopia-fi-forecast

End-to-end data pipeline & forecasting system for Ethiopian financial inclusion (2025–2027). Features dataset enrichment, event-indicator association matrices, hybrid trend models with confidence intervals, and a Streamlit dashboard tracking NFIS 60% access targets. # Ethiopia Financial Inclusion Forecasting System (2025–2027) > **Selam Analytics | 10 Academy Week 11 Challenge** > Autonomous end-to-end Financial Inclusion Forecasting System for Ethiopia (2025–2027) projecting progress toward the National Bank of Ethiopia's (NBE) NFIS-II 60% inclusion target. --- ## 📌 Executive Highlights - **Divergence Paradox Resolved**: Explains why 64M+ registered mobile wallets resulted in only a +3pp net Findex account growth (46% in 2021 to 49% in 2024) due to account duplication, dormancy, and KYC bottlenecks. - **Event-Impact Matrix**: Features an S-curve lag decay propagation model mapping major policies (Fayda e-KYC mandate, EthSwitch QR interoperability, Telebirr launch) to inclusion metrics. - **2025–2027 Scenario Forecasts**: Projecting **56.01% (2025)**, **59.34% (2026)**, and **62.37% (2027)** Account Ownership under the Base Case, achieving the **60% NFIS-II target by 2026/2027**. - **Interactive Web Dashboard**: Built with Streamlit & Plotly featuring KPI metrics, scenario controls, dynamic lag simulator, and CSV export. --- ## 🏗️ Repository Structure ``` ethiopia-fi-forecast/ ├── .github/ │ └── workflows/ │ └── unittests.yml # GitHub Actions CI workflow ├── data/ │ ├── raw/ │ │ ├── ethiopia_fi_unified_data.csv# Starter & baseline Findex observation dataset │ │ └── reference_codes.csv # Indicator reference lookup table │ └── processed/ │ ├── ethiopia_fi_enriched.csv # Macro/event-enriched dataset (35 records) │ └── data_enrichment_log.md # Detailed audit trail of all additions ├── notebooks/ │ ├── 01_data_exploration_and_enrichment.ipynb │ ├── 02_exploratory_data_analysis.ipynb │ ├── 03_event_impact_modeling.ipynb │ └── 04_forecasting_access_and_usage.ipynb ├── src/ │ ├── __init__.py │ ├── data_loader.py # Ingestion, validation, & enrichment engine │ ├── impact_model.py # Association matrix & S-curve lag deca …