All projects
National Winner

KrishiSathi

An AI-powered agricultural advisory platform for farmers, combining crop recommendations, contextual assistance, and a RAG pipeline built around agricultural knowledge.

Problem

Farmers in Nepal lack access to personalized, data-driven agricultural advice. Traditional extension services are limited and slow.

Solution

An AI advisory platform combining RAG pipelines with agricultural datasets and soil information. Uses in-browser ONNX inference to minimize backend costs while delivering fast, relevant recommendations.

Key Features

RAG pipeline with Pinecone vector databases and Groq LLM APIs
In-browser ONNX Runtime inference via Transformers.js
Agricultural dataset and soil information integration
Crop recommendation and farmer assistance

Outcomes

National Level Hackathon Winner (2024)Reduced infrastructure costs via client-side inference

Tech Stack

Next.jsTypeScriptExpressMongoDBPineconeTransformers.jsGroqONNX Runtime

More Projects