Ktor AI Travel Assistant

An AI-powered travel assistant chat application built with Ktor and Kotlin Multiplatform. This application provides personalized travel recommendations, information, and advice through a conversational interface.

Overview

This project demonstrates how to build a modern AI-powered chat application using Kotlin and Ktor. It features:

• A travel agency assistant that provides personalized travel recommendations
• Real-time chat interface using WebSockets and Server-Sent Events (SSE)
• AI-powered responses using OpenAI models via LangChain4j
• Retrieval Augmented Generation (RAG) for enhanced responses with domain-specific knowledge
• Conversation memory to maintain context across messages
• OAuth authentication (with Keycloak)
• Metrics tracking for AI question answering

Technologies

Backend

• Ktor: Kotlin asynchronous web framework
• LangChain4j: Java/Kotlin library for working with large language models
• OpenAI: AI models for generating responses
• Exposed: Kotlin SQL framework for database access
• Flyway: Database migration tool
• Kotlinx.serialization: JSON serialization/deserialization
• WebSockets & SSE: For real-time communication
• Keycloak: For OAuth authentication and authorization

Frontend

• Kotlin Multiplatform: For sharing code between platforms
• Compose Multiplatform: For cross-platform UI development
• Ktor Client: For WebSocket and HTTP communication with the server

DevOps

• Docker: For containerization and deployment
• Gradle: For build automation
• Grafana: For metrics visualization
• PostgreSQL: For data storage
• Nginx: For reverse proxy and routing

Architecture

The application follows a modular architecture:

1. AI Module: Handles interaction with OpenAI models and document retrieval
2. Chat Module: Manages WebSocket/SSE connections and message routing
3. Security Module: Handles OAuth authentication with Keycloak
4. Admin Module: Provides administrative functions for document management
5. ComposeApp Module: Implements the cross-platform user interface

The application uses Retrieval Augmented Generation (RAG) to enhance AI responses with domain-specific knowledge. It can ingest documents to build a knowledge base that the AI can reference when answering questions.

Project Structure

The project follows a multi-module setup:

• app: Main server application with Ktor
• composeApp: Cross-platform UI using Compose Multiplatform
• langchain4j-kotlinx-coroutines: Custom extensions for LangChain4j with Kotlin coroutines support

The project uses a version catalog (see gradle/libs.versions.toml) to declare and version dependencies.

Setup Instructions

Prerequisites

• JDK 17 or higher
• Docker and Docker Compose

Building the Project

./gradlew build

Database Setup

The database is automatically set up when running with Docker Compose. The application uses Flyway to manage database migrations.

Usage Instructions

Running Locally

1. Start the Docker services:

cd docker
docker-compose up -d

2. Run the server:

./gradlew run

Accessing the UI

• The UI is available at http://localhost
• Login with the credentials configured in Keycloak (user: ktor_user, pw: ktor_password)

Example Interactions

You can ask the travel assistant questions like:

• "What are some good destinations for a family vacation in Europe?"
• "I'm planning a trip to Japan in April. What should I know?"
• "What's the best way to travel between Paris and London?"

Deployment Instructions

Docker Setup

The application can be deployed using Docker Compose:

cd docker
docker-compose up -d

This will start:

• PostgreSQL database
• Grafana for monitoring
• Keycloak for authentication
• Nginx as a reverse proxy

Configuration Options

Configuration options can be modified in:

• app/src/main/resources/application.yaml
• docker/docker-compose.yml
• docker/nginx.conf

Development

Testing

• Relies on TestContainers to start Postgres

./gradlew test

Contributing

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Submit a pull request

License

This project is licensed under the Apache 2.0 License - see the LICENSE file for details.