Biyo Dhawr — Water Intelligence & Drought Resilience Platform

Problem Water management in rural Somaliland is reactive: authorities act only after a source dries up. Communities have no reporting channel and no smartphones; government data is scattered across databases, coordination is manual, and decisions are delayed. Solution • Simple reporting: a USSD line (*999#) in Somali lets anyone with a basic phone report a broken or dry water point in seconds. • Live water map: 442 real water points from the SWIMS dataset (Awdal region) on a satellite map with status, water level and maintenance history. • Centralized database: regions → districts → villages → water sources, reports, alerts and sensor readings in one PostgreSQL schema. • AI Intelligence Center: an explainable risk engine scores every village (Low → Severe) and estimates days-to-failure per well; an LLM (Llama via Groq) writes narrative reports with recommended actions. • Real-time dashboard: Socket.IO updates, report triage (verify/reject), analytics and CSV export for government workers and village leaders. My role Full-stack developer: built the dashboard map and filter logic, field-report verification workflow and severity UI, contributed backend endpoints (villages/districts filtering, report severity and action tracking), and open-sourced the project (MIT) with documentation. Team Ibrahim Abdirashid Ibrahim, Abdillahi Abdirahman Abdisalan, Ayub Jama Khalid, Abdulahi.