UgaAI

live

AI farming platform for mandi prices, pest alerts, and scheme eligibility

Next.jsGeminiSupabaseCapacitorvisit ↗

the problem

India has 140M farmers making critical decisions — what to grow, when to sell — based on outdated information or middlemen. Mandi prices exist on government portals but aren't actionable. Pest alerts don't reach farmers in time.

market context

Government mandi portals publish 1,400+ commodity price records daily across 123 commodities and 19 states, but no consumer product makes this data useful for farmers. UgaAI aggregates this, adds AI-powered pest alerts, weather overlays, and scheme eligibility — all in Hindi and English.

how it's built

Next.js with ISR for 100+ SEO pages (commodity + district pages for organic search). Live mandi price data ingested 3× daily via cron. 7 total cron jobs: mandi prices, weather, GDD calculations, pest alerts, government scheme updates. Capacitor wraps the web app into an Android APK.

architecture

tech stack

frontend

Next.jsTailwind CSSLeaflet

backend

SupabaseNext.js API routes

data

Gemini Flash

infra

VercelSupabaseCapacitor (Android)

outcomes

→~85K impressions/month on Google Search
→~580 clicks/month, position 8.1 average
→123 commodities, 19 states covered
→7 automated cron jobs running