Accede Travel Concierge

Disclaimer: This sample is for demonstration purposes only and is not intended for production use. It may not include all the features or safeguards needed for a production environment.

Introduction

Accede Travel Concierge is a modular application designed to streamline travel planning and expense management. The project is structured into three main components:

Project Structure

graph TD
    Service["Backend<br>(Agentic Workflows)"]
    MCPServer["MCPServer"]
    WebUI["Web UI"]
    LocalGuide["Local Guide Agent"]

    %% Relationships
    WebUI -->|Uses APIs| Service
    LocalGuide -->|Provides Data| Service
    Service -->|Interacts With| MCPServer
    MCPServer -->|Processes Requests| Service
    Service -->|Manages| LocalGuide

Loading

.NET Projects

• AccedeSimple.AppHost/: Serves as the entry point and host for the application, managing configuration and startup logic.
• AccedeSimple.Service/: Implements the core business logic and service layer, handling travel planning, approvals, and expense processing.
• AccedeSimple.ServiceDefaults/: Provides default implementations and shared utilities to support the service layer.
• AccedeSimple.Domain/: Contains domain models and logic for features such as approvals, bookings, expenses, trips, and shared utilities.
• AccedeSimple.MCPServer/: Implements the MCP server functionality for extended capabilities.

Other Projects

• webui/: React web application for the user interface.
  • src/components/: Contains React components like AdminPage, ChatContainer, and VirtualizedChatList.
  • src/services/: Includes service files like AdminService.ts and ChatService.ts.
  • src/styles/: Contains CSS files for styling various components.
  • src/types/: TypeScript type definitions for the application.
• localguide/: FastAPI Web API for retrieving city attractions using an AI agent.
  • Dockerfile: Configuration for containerizing the API.
  • main.py: Entry point for the FastAPI application.
  • pyproject.toml: Python project configuration.

Prerequisites

To run the application, ensure the following tools and frameworks are installed:

• .NET 9 SDK or greater
• Python 3.12 or greater
• UV
• Visual Studio Code
• Azure AI Foundry Project within the above subscription and resource group (and within an AI Foundry Hub) that can be used for content safety evaluations. See Setting Up Azure AI Foundry for Safety Evaluations.
  • Azure OpenAI Resource with the gpt-4.1 and text-embedding-3-small model deployed and permissions.
• Azure Developer CLI (if applicable)

Quick Start

1. Clone the repository:

   git clone https://github.com/your-repo/AccedeSimple.git
   cd AccedeSimple

2. Install dependencies: Ensure you have the required .NET SDK installed, then restore dependencies:

   .NET

   dotnet restore

   Python

   cd src/localguide
   uv sync

3. Configure user secrets

   1. Navigate to the _src/AccedeSimple.AppHost project.
   2. Set the following user secrets. i.e. dotnet user-secrets set "AzureOpenAI:ResourceGroup" "YOUR-VALUE"
      • AzureOpenAI:ResourceGroup - The name of your Azure Resource Group where the OpenAI Resource is deployed to
      • AzureOpenAI:ResourceName - The name of your Azure OpenAI Resource
      • AzureOpenAI:Endpoint - The endpoint for youor Azure OpenAI Resource
      • Azure:SubscriptionId - The subscription ID you deployed your resources to
      • Azure:ResourceGroup - The name of your Azure OpenAI Resource is deployed to.
      • Azure:Location - The location you deployed your Azure OpenAI Resource to.
      • Azure:AllowResourceGroupCreation - Set to false to use existing resource.
      • AzureAIFoundry:Project - The name of the Azure AI Foundry Project.

Running the app

1. Run the application:

   Start the application using the .NET CLI:

   dotnet run --project src/AccedeSimple.AppHost

You're now ready to use the Accede Travel Concierge application!

Deployment

Follow the standard deployment guidance for Aspire

1. In the root directory, run the following command

   azd init

2. Deploy the app

   azd up