🌱 AgroAI - Intelligent Plant & Animal Health Monitoring Platform

AgroAI is a highly polished, responsive, and beautiful AI-powered diagnostic platform designed for modern agriculture and animal husbandry. It provides end-to-end intelligent vision pipelines for real-time plant crop pathology, insect infestation analysis, animal health diagnostics, nearby veterinary care integration, local history tracking, and professional agricultural report generation.

🚀 Live Production Link: https://agroai-kappa.vercel.app

---

✨ Features & Capabilities

1. 🔬 Upgraded AI Vision Diagnostics

• Advanced SOTA VLM Engine: Powered by Qwen/Qwen3-VL-235B-A22B-Instruct—a 235-billion-parameter state-of-the-art vision-language model. This provides exceptional accuracy for identifying plant species, crop diseases, pest damage, and nutrient deficiencies.
• Dual Processing Pipeline:
  • Single Model Mode: Direct multimodal vision analysis utilizing Qwen3-VL-235B for lightning-fast, high-context agricultural reasoning.
  • Multi-Model Mode: Sequential execution utilizing YOLOv11 (leaf localization & cropping) ➡️ ViT Classifier (species prediction) ➡️ Llama 3.1 8B Instruct (actionable recommendation writing).

2. 📸 Native Mobile Camera & Gallery Integration

• Refactored file upload systems to include custom styled side-by-side controls:
  • 📁 Choose Image: Allows standard local photo uploads from desktop or mobile libraries.
  • 📷 Take Photo: Hooks directly into mobile hardware to launch the rear (environment) camera instantly (capture="environment"), making crop field analysis seamless.

3. 🌐 End-to-End Multilingual Support (18 Global Languages)

• Integrated native translation capabilities across:
  • Speech Recognition: Live voice-to-text recording (Whisper/Web Speech API) with language selection.
  • Contextual Prompt Injection: Injects translation commands dynamically based on selected language codes (e.g. hi-IN for Hindi, bn-IN for Bengali, es-ES for Spanish) so that the entire diagnosis, symptoms, and treatments are generated natively in that language by the AI!
  • Text-to-Speech (TTS): Read Aloud synthesis for audio feedback of treatment recommendations in the selected language.

4. 🐾 Animal Health Diagnostic Pipeline & Map Finder

• Seamlessly analyzes medical conditions, injuries, and severity levels in livestock and household pets.
• Veterinary Hospital Finder: Automatically queries location APIs to discover nearby vet clinics and maps them onto a custom Leaflet map interface when moderate/severe symptoms are detected.

5. 🗄️ Serverless Ready Cloud Storage (Vercel & MongoDB Atlas)

• Configured serverless execution for production environments using Vercel.
• Dynamic Writable Path Router: Checks runtime environments and utilizes the system's write-enabled /tmp directory under read-only serverless constraints (resolving file-system startup crashes).
• Integrates securely with MongoDB Atlas database cloud storage for persistent diagnostics history and stats.

---

🛠️ Tech Stack

• Frontend: Vanilla HTML5, Vanilla CSS3 (Custom styling grid & dark/light theme), JavaScript (ES6+), Leaflet Map & CartoDB tiles.
• Backend: Flask (Python), PyMongo, MongoDB Atlas, Pillow, OpenAI API Client, certifi (SSL).
• Deployment: Vercel Serverless Functions.

---

🚀 Getting Started Locally

1. Prerequisites

• Python 3.10+
• Hugging Face Account (for API tokens)
• MongoDB Atlas Cloud Account (or runs on automatic local JSON database fallback)

2. Installation & Setup

1. Clone the repository:

   git clone https://github.com/Rohan-R07/agroAI.git
   cd agroAI

2. Install dependencies:

   pip install -r backend/requirements.txt

3. Configure Environment Variables: Create a local configuration folder in backend/data/ or set environment variables in your local environment:

   • MONGO_URI = your_mongodb_connection_string
   • HF_TOKEN = your_huggingface_api_token

4. Run the Application:

   python backend/app.py

   Open your browser and navigate to http://localhost:5000 to interact with the system locally.

---

☁️ Cloud Deployment (Vercel CLI)

1. Install Vercel CLI:

   npm install -g vercel

2. Link the Project:

   vercel link --scope anik-da-projects1 --project agroai --token YOUR_TOKEN --yes

3. Configure Environment Settings:

   vercel env add MONGO_URI production --value "YOUR_MONGO_URI" --token YOUR_TOKEN --yes
   vercel env add HF_TOKEN production --value "YOUR_HF_TOKEN" --token YOUR_TOKEN --yes

4. Deploy to Production:

   vercel --prod --token YOUR_TOKEN --yes

---

🔒 Security & Standards

• Dynamic environment variable resolution.
• Sensitive local diagnostic photos and access keys are ignored in .gitignore.
• Production Vercel routes use SSL-certified pipelines.