Article’s

Operational information fragmentation in agriculture remains a significant barrier to productivity, crop yield protection, and economic security for smallholder farmers. Traditional farming methods rely heavily on static calendars and localized, often asymmetrical, information regarding soil health, disease diagnostics, weather patterns, and market pricing. This paper presents AgriSathi, a unified, high-performance web portal designed to address these gaps. Implemented with a FastAPI backend and a reverse-proxied Nginx server, AgriSathi integrates eight core agricultural modules: crop recommendations based on multi-parameter environmental rules, non-invasive leaf disease diagnostics with confidence thresholds exceeding 85%, localized weather forecasting with tailored agronomic advisories, real-time wholesale price tracking across multiple weight classes, and a stateful multilingual voice assistant supporting regional languages (specifically Telugu). The system architecture leverages SQLite for low-latency transaction processing. Preliminary validation demonstrates that the unified interface reduces information retrieval latency and empowers farmers with actionable, data-driven decisions to prevent yield loss and middleman exploitation. Keywords: Precision Agriculture, FastAPI, Convolutional Neural Networks, Multilingual Localization, Leaf Pathology, Mandi Price Index, Agricultural Decision Support Systems.