ai-lc4j-demos/README.md

50 lines
No EOL
2.4 KiB
Markdown

kvant AI Model as a Service & LangChain4j
===============================================
This demos helps you get started with the integration of AI in Java based applications using Quarkus and LangChain4j.
Java is a widely used language established over the last 2 decades. There are a huge number of applications developed
using Java, including many deployed in enterprise environments. This makes it appropriate to approach AI from
the Java world.
# Goals
You are going to play (and hopefully enjoy) with:
* How to integrate LLMs (Language Models) in your Java application
* How to build a chatbot using LangChain4j
* How to configure and how to pass prompts to the LLM
* How to build agentic systems
* How to build simple and advanced RAG (Retrieval-Augmented Generation) patterns
* How to deal with prompt injection: Guardrails
* How to look in the room-machine: Observability
# Requirements
* Acquire kvant AI Model as a Service tokens. You can start with the FREE plans!!! - https://console.kvant.cloud/en/products/categories/CLOUD_MAAS
* Java SDK (version 21.x.x, recommend GraalVM CE) - https://www.graalvm.org/downloads/#
* Apache Maven JDK (version >= 3.9.x) - https://maven.apache.org/download.cgi
* Git (version >= 2.34.x) - https://git-scm.com/downloads
* Docker Desktop - https://www.docker.com/products/docker-desktop/ (version >= 4.35.x)
* LLM Access tokens - They will be provided during the workshop
* (Recommended) IntelliJ IDEA - https://www.jetbrains.com/idea/download/ (version >= 2024.2.5)
# Run demos
1) Clone the `ai-lc4j-workshop` repository - https://git.kvant.cloud/phoenix/ai-lc4j-demos
2) Move to your `ai-lc4j-workshop` folder
3) Move to the corresponding `demo-0x` folder
4) Execute `mvn clean quarkus:dev`
5) Open your internet browser a point to `http:\\localhost:8080`
**WARNING:** Be sure your Docker environment is running
# Credits
These demos are based on the existing LangChain4j examples and workshops. They have been adapted to use
Phoenix Technologies AG's kvant AI cloud platform and LLM's hosted on it.
# Resources
* Phoenix Technologies AG - https://phoenix-technologies.ch/en
* Kvant AI Cloud - https://kvant.cloud/en/products/kvant-ai/ai-concierge
* LangChain - https://www.langchain.com/
* LangChain4j - https://docs.langchain4j.dev/intro/
* LangChain4j Tutorials - https://docs.langchain4j.dev/category/tutorials
* Quarkus LangChain4j - https://docs.quarkiverse.io/quarkus-langchain4j/dev/index.html
* Quarkus - https://quarkus.io/