Back to Projects

Team Project — March 2026

AGRITECH

AI-Powered Smart Agriculture Platform

Overview

Built a MERN platform for AI-driven smart agriculture, enabling semi-literate farmers to interact through voice commands, Text-to-Speech (TTS) and image-based crop analysis.

Technology Stack

Next.jsTypeScriptTailwind CSSNode.jsExpress.jsMongoDBFramer MotionPython AI

Key Highlights

  • Project Concept: Designed an inclusive, AI-driven agronomy platform tailored for semi-literate farmers, enabling seamless interaction through voice commands, Text-to-Speech (TTS) and image-based crop analysis.
  • AI & Disease Detection: Trained a custom Python AI model using a dataset of 9,000+ images for precise crop disease identification, complemented by GoogleAI Studio for generating actionable, multi-lingual remedies.
  • Core Smart Features: Integrated deep-dive soil intelligence diagnostics, localized live weather insights, automated crop monitoring and precision "guess farming" forecasting modules.
  • Architecture & Scalability: Engineered the entire system architecture using Next.js (ISR,SSR,CSR), Tailwind CSS, TypeScript, Node.js, Express.js and MongoDB, delivering highly optimized and type-safe backend services.
  • UI/UX & Deployment: Built a highly accessible, premium user interface featuring smooth transitions with Framer Motion; automated the workflow using a robust CI/CD pipeline deployed on Vercel and able to host a self-hosted VPS.

Challenges Faced

  • Training a custom Python model on 9,000+ images and keeping real-time crop disease detection fast enough for a production web flow.
  • Designing voice-command and Text-to-Speech flows that work reliably for semi-literate users across multiple languages on low-end devices.
  • Balancing SSR, ISR and CSR rendering modes in Next.js while keeping the backend type-safe and performant for a large team codebase.
  • Managing a consistent CI/CD pipeline across Vercel and a self-hosted VPS with environment-specific configuration.

Improvements & Future Plans

  • Expand multilingual support and add offline-first voice responses for remote areas.
  • Launch a companion mobile app (React Native / Expo) with offline disease detection.
  • Add more agronomy models such as yield prediction and pest detection, with model versioning.
  • Introduce a community marketplace, push-notification weather alerts and real-time sensor integrations.