shopping-assistant-api

.NET Back-end API with OpenAI integration for a Shopping Assistant that utilizes Natural Language Processing (NLP) technology to interpret user queries for products and gifts. Users interact through a chat-style interface to communicate their shopping requirements.

Table of Contents

• Features
• Stack
• Installation
  • Prerequisites
  • Setup Instructions
• Configuration

Features

• Chat-style interface that processes natural language queries for product and gift recommendations.
• Integration with OpenAI API to generate intelligent product search responses.
• Server-Sent Events (SSE) streaming product search results to clients in real-time.
• Personal wishlist management with CRUD operations.
• Role and user management with authorization and JWT-based authentication.
• GraphQL API implemented with HotChocolate for flexible queries and mutations.
• Pagination support for roles, users, wishlists, messages, and products.

Stack

• Language: C# (.NET 7)
• Frameworks & Libraries:
  • ASP.NET Core Web API
  • HotChocolate GraphQL
  • MongoDB with official C# driver
• Authentication & Security:
  • JWT Bearer Authentication
  • Password hashing with PBKDF2 (Rfc2898DeriveBytes)
• Cloud & DevOps:
  • Azure App Configuration
  • GitHub Actions for CI/CD workflows
• External APIs:
  • OpenAI API for natural language processing and chat completions
• Tools & Extensions:
  • Swagger / OpenAPI for API documentation
  • Server-Sent Events for realtime streaming responses
  • Visual Studio Code DevContainer for development environment

Installation

Prerequisites

• .NET 7 SDK
• MongoDB instance (local or hosted)
• An OpenAI API key to access the OpenAI services

Setup Instructions

1. Clone the repository:

   git clone https://github.com/Shchoholiev/shopping-assistant-api.git
   cd shopping-assistant-api
   

2. Navigate to the API project and restore dependencies:

   dotnet restore ShoppingAssistantApi.Api/ShoppingAssistantApi.Api.csproj
   

3. Build the solution:

   dotnet build ShoppingAssistantApi.sln
   

4. Configure the environment variables (see Configuration section).

5. Run the API locally:

   cd ShoppingAssistantApi.Api
   dotnet run
   

6. The API will start and Swagger UI will be available (usually at https://localhost:7268/swagger).

7. Use GraphQL Playground or REST clients to interact with the API.

Configuration

The application uses configuration files and environment variables to configure key settings. Important configuration values are:

• MongoDB connection:

  • Set your MongoDB connection string (default database name is ShoppingAssistant) in appsettings.json or environment variable:

    "ConnectionStrings": {
      "MongoDatabaseName": "ShoppingAssistant"
    }
    

  • The connection string URL is typically configured externally.

• JWT Authentication: Configure JWT validation parameters in your configuration with keys like:

  "JsonWebTokenKeys": {
    "ValidateIssuer": true,
    "ValidateAudience": true,
    "ValidateLifetime": true,
    "ValidateIssuerSigningKey": true,
    "ValidIssuer": "<your-issuer>",
    "ValidAudience": "<your-audience>",
    "IssuerSigningKey": "<your-secret-key>"
  }
  

• OpenAI API: API endpoint and API key settings:

  "OpenAi": {
    "ApiUrl": "https://api.openai.com/v1/chat/completions",
    "ApiKey": "<your_openai_api_key>"
  }
  

  Set the API key securely via environment variables or Azure App Configuration.