# java-ai-playground

**Repository Path**: jerrycell/java-ai-playground

## Basic Information

- **Project Name**: java-ai-playground
- **Description**: No description available
- **Primary Language**: Unknown
- **License**: MIT
- **Default Branch**: dependabot/maven/quarkus/com.github.mcollovati-quarkus-hilla-24.6.0
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2025-06-19
- **Last Updated**: 2025-06-19

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# Java AI library comparison app

![A browser window with a chatbot and a data grid displaying flight booking details](screenshot.jpg)

This app is an AI-powered customer support application that:

- Has access to terms and conditions (retrieval augmented generation, RAG)
- Can access tools (Java methods) to perform actions
- Uses an LLM to interact with the user

The application includes implementations for: 

- [LangChain4j](https://github.com/langchain4j/langchain4j) in the `main` branch
- [Quarkus](https://quarkus.io) in the `quarkus` branch (thanks to [@edeandrea](https://github.com/edeandrea)!)
- [Spring AI](https://spring.io/projects/spring-ai/) in the `spring-ai` branch (thanks to [@tzolov](https://github.com/tzolov)!)
- [Semantic Kernel](https://github.com/microsoft/semantic-kernel) in the `semantic-kernel` branch (thanks to [@sohamda](https://github.com/sohamda)!)

The UI is built using [Vaadin Hilla](https://vaadin.com) and the backend is built using Quarkus.

## Requirements
- Java 17+
- OpenAI API key in `OPENAI_API_KEY` environment variable (optional if you want to use OpenAI)

## Running
Run the app by running `./mvnw quarkus:dev`. By default it will communicate with OpenAI.

If you have [Ollama](https://ollama.com/) installed and would prefer to use that instead, run `./mvnw quarkus:dev -Dquarkus.profile=ollama`.

## Thanks
This demo was inspired by the [LangChain4jCustomer Support Agent example](https://github.com/langchain4j/langchain4j-examples/tree/main/spring-boot-example/src/main/java/dev/langchain4j/example).

I want to thank the LangChain4j, Quarkus, Spring AI, and Microsoft teams for their support in building this demo.
Especially, I want to thank [@edeandrea](https://github.com/edeandrea) from the Quarkus team for his help in building the Quarkus implementation, [@tzolov](https://github.com/tzolov) from The Spring AI team for his help in building the Spring AI implementation, and [@sohamda](https://github.com/sohamda) from Microsoft for the Semantic Kernel implementation.