Logo Lanfrica

hajisml/niwapi

Domain:

climateenvironment and energy

Record type:

software
Creator:
haj
Host:
NiWapi (Swahili for "Where is it?") is a two-way climate risk intelligence and early-warning platform designed for the Lake Victoria Basin and high-risk Kenyan regions. It bridges the gap between high-altitude satellite data and on-the-ground human reality. # NiWapi **Climate Risk & Proactive Drainage Resilience Platform** Built for the **Zone01 Kisumu GreenTech Hackathon 2026**, Track 6: Climate Risk Intelligence & Resilience Platform. ## Live Demo - **Citizen field app:** niwapi.vercel.app - **Municipal dashboard:** niwapi-municipal.vercel.app - **Backend API:** niwapi-backend.onrender.com (docs at `/docs`) The backend is on Render's free tier, which spins down after ~15 minutes idle — the first request after a gap can take 20-30s to wake up. ## The Problem Urban and peri-urban centers across the Lake Victoria Basin face recurrent flash floods driven by extreme weather events and unmanaged surface runoff. While regional weather warning systems exist, urban flood vulnerability is heavily aggravated by local drainage failures (e.g., stormwater channels blocked by solid waste, siltation). Remediation is traditionally reactive—teams deploy *after* damage occurs. ## The Solution **NiWapi** is an end-to-end climate risk intelligence platform that shifts flood response from reactive disaster management to proactive infrastructure readiness. It merges crowdsourced citizen blockage reports, real-time IoT water-level sensors, and predictive weather modeling to map, score, and remediate drainage chokepoints *before* rainfall triggers floods. ## System Architecture The platform consists of several key components: - **Citizen App (PWA):** A React/Vite-based Progressive Web App allowing users to capture geo-tagged photos of drainage blockages. Features offline caching (`IndexedDB`) to support low-connectivity field reporting. - **FastAPI Backend:** Orchestrates data flow, linking images to Gemini Vision API for severity classification, fetching rainfall data from Open-Meteo, and calculating localized composite risk indices. - **Supabase Layer:** Centralized spatial database utilizing PostgreSQL and PostGIS to query map data (`geometry(Point, 4326)`), store incident images, and manage real-time work …