Nigehbaan

A dropout-risk early warning system for Pakistani girls' schools, built for AI Seekho Builders Day 2026 at NIC Islamabad. A deterministic Dart rule engine scores attendance, assessment, participation and fee signals, then applies a co-occurrence multiplier when several different families of signal degrade in the same window, because that overlap is the reliable warning sign rather than any single lapse.

Gemini 2.5 Flash reads those trends alongside the teacher's own written observations, and is constrained so it can never compute the score itself and must cite specific evidence before it may raise or lower a verdict. It drafts a respectful Urdu message home that the school sends, rather than the app messaging a family on its own. Open sourced under MIT with 20 screens and 47 automated tests.