Archive · 9 Sep 2026
IIT-Kanpur AI Family Score Could Transform Welfare Delivery
IIT-Kanpur has created a computer system called Family Score.
It helps governments find families that may need welfare assistance.
The system studies information the government already has.
This means officials may not need to conduct new surveys for every household.
The system uses about 15 government datasets to estimate a family’s socio-economic condition.
Its results were compared with scores from a statewide survey.
IIT-Kanpur said the AI models performed better than conventional surveys.
The project was developed with the Andhra Pradesh government and is intended to make welfare delivery more targeted and transparent.
IIT-Kanpur developed an AI-powered Family Score system to identify households needing welfare support.
The system uses data from about 15 government-owned datasets instead of requiring fresh household surveys.
AI models classify families by socio-economic conditions and can generate dynamic scores from available records.
IIT-Kanpur said the models outperformed conventional surveys while requiring less cost and effort.
The initiative was developed with the Andhra Pradesh government to support transparent and responsible AI in public services.
- Who
- The Indian Institute of Technology-Kanpur, working with the Andhra Pradesh government, developed the Family Score system.
- What
- An AI system that assesses household socio-economic conditions and helps identify potential welfare beneficiaries.
- When
- The findings were presented on Tuesday; the articles do not specify the date.
- Where
- The initiative was conducted in collaboration with the Government of Andhra Pradesh.
- Why
- To identify deserving households more quickly and cost-effectively using existing government data, while reducing reliance on fresh surveys.
- System
- Family Score, an AI-powered household assessment system
- Developers
- Indian Institute of Technology-Kanpur in collaboration with the Andhra Pradesh government
- Data used
- About 15 government-owned datasets containing household socio-economic indicators
- Benchmark
- The models were calibrated against socio-economic scores from a statewide survey
- Reported performance
- IIT-Kanpur said its ensemble of AI models outperformed conventional surveys
- Potential benefit
- Faster, lower-cost identification of households for targeted welfare support
- Presentation
- The findings were presented to Andhra Pradesh Chief Minister N Chandrababu Naidu and senior officials