StoreSmart
AgTech · Predictive ML · IoT · India · Indian Patent · 2022
India loses 30 to 40% of its produce to post-harvest storage failures every year. The cold storage infrastructure exists. What doesn't exist is intelligence: farmers and cooperative managers have no early warning before spoilage begins, and by the time it's visible it is already too late.
The product was designed entirely around the conditions of rural Karnataka: intermittent connectivity, limited smartphone literacy, cooperative managers responsible for dozens of farmers simultaneously. Every technical decision had to survive that context.
The alert system was SMS-first - not because it was easier to build, but because it was the only channel guaranteed to reach the user. The dashboard was built for weekly review cycles, not hourly monitoring. Sensor hardware was designed for installation by non-technical staff in under 20 minutes, because a system that requires an engineer to set up will never scale in this market.
The 24 to 48 hour prediction window was the core product bet: early enough to act, specific enough to be useful. A vaguer warning further out would be ignored.
