VeeThoithi/DemographIQ

AI-powered population forecasting and planning intelligence system for Kenya # DemographIQ **AI-Powered Population Forecasting & Planning Intelligence Platform for Kenya** --- ## Problem Statement Kenya's public sector planners -- at both national and county government levels -- face a persistent challenge: population data is only available at 10-year census intervals (KNBS 2019 being the most recent), yet infrastructure, healthcare, and education planning decisions must be made annually. Existing tools either operate at too coarse a national level, require expensive proprietary software, or are inaccessible to county-level planners without specialist data science skills. DemographIQ addresses this gap by building an open, full-stack demographic intelligence platform that: - Interpolates and forecasts Kenya's national population from 1962 to 2035 using two independent ML approaches - Disaggregates national forecasts to all 47 counties - Translates population growth data into actionable planning recommendations (healthcare, education, infrastructure, water) through a publicly accessible web dashboard --- ## Overview DemographIQ is a demographic intelligence system designed to forecast population growth and support data-driven planning across Kenya. The platform combines machine learning, data visualization, and rule-based planning intelligence to provide both **national and county-level population predictions**, along with **automated infrastructure and policy recommendations**. --- ## Data Sources All demographic data used in this project is sourced from publicly available official statistics: - **Kenya National Bureau of Statistics (KNBS)** -- 2019 Kenya Population and Housing Census - Volume I: Population by County and Sub-County - Volume II: Distribution of Population by Administrative Units - **World Bank Open Data** -- National-level indicators including GDP per capita, urban population percentage - **KNBS / WHO** -- Health indicators including fertility rate and infant mortality rate > Full column-level documentation is a …