Devoxx Greece 2024 - Quarkus meets AI : Build your own LLM-powered application

Sergey Kolosov
Ai

Discover how Quarkus and Linkchain4j can be used to build powerful AI applications with LLMs, including enhanced chatbots, intelligent customer service, and data analysis.

Key takeaways
  • LLMs (Language Model LARGE Models) are a type of AI model that can:
    • Generate human-like text based on input prompts
    • Answer questions and respond to user messages
    • Learn from vast amounts of text data
  • Linkchain4j, a Java library, can integrate with LLMs and other AI models
  • Devoxx Conference Talk: “Quarkus meets AI: Build your own LLM-powered application”
  • The talk covers:
    • Introduction to LLMs and their capabilities
    • Using Linkchain4j to create an LLM-powered application
    • Quarkus and its role in integrating with LLMs
    • Demo examples showing LLMs in action
  • LLMs can be used for:
    • Enhanced chatbots
    • Intelligent customer service
    • Data analysis and interpretation
    • Content creation
  • Fine-tuning LLMs is crucial for adapting them to specific use cases
  • Quarkus provides an easy-to-use interface for working with LLMs
  • The talk showcases how to create an LLM-powered application using Quarkus and Linkchain4j
  • Real-world examples of LLMs include:
    • AlphaZeroGo (Go-playing AI)
    • AlphaZeroChess (chess-playing AI)
    • ChatGPT (AI chatbot)
  • LLMs can be used for creative tasks such as:
    • Generating music
    • Creating art
    • Writing essays
    • Playing games
  • The speaker emphasizes the importance of:
    • Using LLMs responsibly
    • Ensuring fairness and transparency in AI development
    • Considering the ethical implications of AI
  • The talk concludes with a demonstration of LLMs in action using Quarkus and Linkchain4j.