Recommendation Website with AI Chatbot

This project is a recommendation website enhanced with an AI chatbot, designed for scalability, maintainability, and flexibility using Clean Architecture principles.

Features

• AI Chatbot: Provides personalized recommendations for users, enhancing their experience.
• Clean Architecture: Ensures a well-structured, modular application that's easy to maintain and scale as your project grows.
• Image Management: Leverages Cloudinary for efficient handling and storage of images.
• Google OAuth: Implements secure user authentication using Passport.js and Google OAuth.
• Geminie Vision Integration: Seamlessly integrates Geminie Vision to extract restaurant menu data and generate JSON files for easy ingestion.
• Rasa Chatbot: Utilizes Rasa, a powerful AI framework, to run the chatbot on a separate server for optimized performance.

Technologies Used

• Frontend: Vuejs / Typescript
• Backend: Node.js, Express.js , TypeScript
• Database: MongoDB
• Authentication: Google OAuth with Passport.js
• Image Management: Cloudinary
• AI Chatbot: Rasa
• Image Processing: Geminie Vision

Project Structure

    src/
      ├── Application/ (interfaces and core logic)
      ├── Domain/ (entities)
      ├── Infrastructure/ (usecases/ repository implementation & database or any external system implementation that depends on the interfaces)
      └── Main/ (factories and endpoints)

Explanation of Clean Architecture Folders:

• Application: This folder contains interfaces and core logic for your application. It defines what the system should do without relying on specific technologies or implementations.
• Domain: This folder houses your domain entities, representing the core concepts and rules of your recommendation system.
• Infrastructure: This folder holds concrete implementations of use cases and repositories, along with any external system integrations like Cloudinary or Rasa. It depends on the interfaces defined in the Application layer.
• Main: This folder serves as the entry point, containing factories to create concrete instances and endpoints for your API.

Getting Started

Before diving in, ensure you have the following prerequisites installed:

• Node.js
• MongoDB
• A Cloudinary account
• Google OAuth credentials
• A running Rasa server