Aastha92087/financial-inclusion-pipeline

Data pipeline + ML analysis of financial inclusion in East Africa, using Snowflake, dbt, and Python. # Financial Inclusion Analysis — East Africa An end-to-end data pipeline and analysis project exploring financial inclusion across Kenya, Rwanda, Tanzania, and Uganda, using demographic survey data. ## Overview This project takes raw survey data from Kaggle and builds a complete pipeline from ingestion to a governed data warehouse, through business-level SQL modeling, into Python-based machine learning and an interactive dashboard to answer a real question: who is financially excluded in East Africa, and why? ## Pipeline Architecture ``` Kaggle CSV (raw survey data) │ ▼ Snowflake (loaded via dbt seed) │ ▼ dbt staging model → cleaned column names, filtered nulls │ ▼ dbt mart model → aggregated business metrics (bank access rate by segment) │ ▼ Python (pandas, scikit-learn, matplotlib, plotly) │ ▼ Interactive dashboard (Streamlit) ``` Tools used: Snowflake, dbt, SQL, Python (pandas, scikit-learn, matplotlib, plotly, Streamlit), Git/GitHub ## Dataset Financial inclusion survey data covering ~23,500 respondents across Kenya, Rwanda, Tanzania, and Uganda, including demographics (age, gender, education, job type, household size, location) and whether each respondent has a bank account. ## Business Question Which demographic groups have the least access to formal banking, and what factors predict or explain bank account ownership? ## 1. Data Modeling (dbt + Snowflake) `stg_financial` staging model: cleaned, snake_case column names, filtered null IDs `bank_access_summary` mart model: bank account ownership rate grouped by country, education, job type, and gender, filtered to groups with 30+ respondents to avoid small-sample noise ## 2. Exploratory Finding Women with no formal education, working in informal, self-employed, or no-income categories, have bank account access rates below 3% - consistent across four countries and sample sizes as large as 1,383 people. This points to gender and education as compounding barriers to financial inclusion, not indepe …