🚀 Spring AI Comprehensive Guide

---

"The future of AI is not just about making machines smarter, but about making them more accessible and integrated into our everyday applications." — Unknown

📖 Overview

Spring AI Guide is a comprehensive educational project that demystifies the integration of Large Language Models (LLMs) into Spring Boot applications. This repository serves as a complete guide to understanding and implementing AI-powered features using Spring AI framework, covering everything from basic chat implementations to advanced Retrieval-Augmented Generation (RAG) systems, Model Context Protocol (MCP), and intelligent tool calling.

Whether you're a beginner looking to understand how LLMs work internally or an experienced developer wanting to leverage Spring AI's powerful abstractions, this guide provides hands-on examples with real-world implementations using OpenAI, Ollama, and vector databases.

---

✨ Features

🤖 ChatClient Implementation - Seamless integration with LLM providers (OpenAI, Ollama) for conversational AI
⚡ High-Performance Responses - Optimized implementations for fast model responses
📝 Dynamic Prompt Templates - Variable injection and template-based prompt engineering
🗄️ Vector Store Integration - Embed and store data in PgVector for semantic search
🔍 Similarity Search - Intelligent retrieval based on semantic similarity
🧠 RAG Implementation - Complete Retrieval-Augmented Generation system for context-aware responses
💾 Long-Term Memory - VectorStoreChatMemoryAdvisor for persistent conversation context
🎯 Short-Term Memory - MessageChatMemoryAdvisor for session-based chat history
💬 Hybrid Memory Strategy - Intelligent fallback between memory layers
📊 Database-Backed Chat History - Persistent storage and lookup of conversation data
❓ QuestionAnswer Advisor - Specialized advisor optimized for RAG-based Q&A systems
🛡️ SafeGuard Advisor - Policy enforcement and governance to prevent irrelevant conversations
📈 Custom Token Usage Tracking - Monitor and optimize LLM token consumption
🔧 Tool Calling - External API integration (OpenWeatherMap) with function calling
🐳 Dockerized Infrastructure - Easy setup with Docker for PgVector and dependencies

---

🛠️ Technologies

Tech Stack

Technology	Version	Purpose
Spring Boot	3.x	Application framework and dependency injection
Spring AI	Latest	AI integration framework for LLMs and vector stores
Java	17+	Primary programming language
PgVector	Latest	PostgreSQL extension for vector similarity search
Docker	Latest	Containerization for database services
OpenAI API	GPT-4o-mini	Cloud-based LLM provider
Ollama	Latest	Local LLM runtime
PostgreSQL	Latest	Relational database with vector support
DBeaver	Latest	Database management and visualization
Postman	Latest	API testing and development
Lombok	Latest	Java boilerplate code reduction

---

📸 Screenshots

1. ChatClient Successfully Implemented - Basic ChatClient integration with Spring AI framework

---

2. Fast Response with OpenAI - Performance demonstration with OpenAI models showing rapid response times

---

3. PromptTemplate Variables -Dynamic variable injection into prompts for flexible prompt engineering

---

4. Vector Store Embedding - Data storage in embedded format using PgVector database

---

5. Similarity Search - Semantic search from vector store based on input text queries

---

6. RAG Implementation - Custom vector store search integration for Retrieval-Augmented Generation

---

7. VectorStoreChatMemoryAdvisor - Long-term conversation memory persistence using vector storage

---

8. VectorMemory Response - LLM answers using GPT-4o-mini model with vector memory context

---

9. MessageChatMemoryAdvisor - Short-term session memory implementation for active conversations

---

10. Hybrid Memory Response - Intelligent fallback between memory layers (MessageChatMemory → VectorStore)

---

11. Chat Memory in Database - Persistent conversation storage and lookup in database

---

12. QuestionAnswer Advisor - Specialized advisor optimized for RAG-based Q&A systems

---

13. SafeGuard Advisor - Policy enforcement and governance to prevent irrelevant conversations

---

14. Custom Token Usage Advisor - Token tracking and usage monitoring for cost optimization

---

15. Tool Calling with OpenWeatherMap - External API integration using function calling capabilities

---

---

🚀 Installation

Prerequisites

• Java 17 or higher
• Docker and Docker Compose
• Maven 3.8+
• OpenAI API Key (or Ollama for local models)
• Git

Steps

1. Clone the repository

   git clone https://github.com/ARONAGENT/Spring_AI.git
   cd Spring_AI

2. Set up PgVector with Docker

   docker-compose up -d

3. Configure environment variables

   Create an application.properties or application.yml file:

   spring.ai.openai.api-key=${OPENAI_API_KEY}
   spring.datasource.url=jdbc:postgresql://localhost:5432/vectordb
   spring.datasource.username=postgres
   spring.datasource.password=postgres

4. Build the project

   mvn clean install

5. Run the application

   mvn spring-boot:run

6. Access the application

   The application will be available at http://localhost:8080

---

💡 Usage

Basic ChatClient Example

@Autowired
private ChatClient chatClient;

public String chat(String userMessage) {
    return chatClient.prompt()
            .user(userMessage)
            .call()
            .content();
}

PromptTemplate with Variables

PromptTemplate promptTemplate = new PromptTemplate(
    "Tell me about {topic} in {style} style"
);
Map<String, Object> model = Map.of(
    "topic", "Spring AI",
    "style", "professional"
);
Prompt prompt = promptTemplate.create(model);

Vector Store and RAG

// Store documents
vectorStore.add(List.of(
    new Document("Spring AI is a framework...", metadata)
));

// Similarity search
List<Document> results = vectorStore.similaritySearch(
    SearchRequest.query("What is Spring AI?").withTopK(5)
);

Tool Calling

@Bean
public FunctionCallback weatherFunction() {
    return FunctionCallback.builder()
        .function("getCurrentWeather", this::getWeather)
        .description("Get current weather for a location")
        .build();
}

For detailed examples and implementation guides, please refer to the source code and inline documentation.

---

🤝 Contributing

Contributions are welcome! However, since this is a proprietary educational project, please follow these guidelines:

1. Fork the repository for personal learning and experimentation
2. Open an issue to discuss proposed changes before submitting
3. Submit pull requests with detailed descriptions of your improvements
4. Follow the code style and conventions used in the project
5. Add tests for new features when applicable

Please note that all contributions will be subject to the project's proprietary license.

---

📄 License

Copyright (c) 2024 ROHAN UKE

This project and its source code are the exclusive property of the author.
**Unauthorized copying, modification, distribution, or commercial use is strictly prohibited.**
Limited use is granted for learning, reviewing, and non-commercial demonstration purposes only.
No warranties are provided; use at your own risk.

For permissions beyond this notice, contact: rohanuke1@example.com

---

🙏 Acknowledgments

This project was made possible thanks to:

• Spring AI Team at Pivotal for creating an excellent framework for AI integration
• OpenAI for providing powerful and accessible LLM APIs
• PgVector maintainers for the PostgreSQL vector similarity search extension
• Ollama team for making local LLM deployment simple and efficient
• The open-source community for continuous innovation and knowledge sharing
• Docker for simplifying development environment setup
• All contributors who have helped improve this educational resource

Special thanks to the broader Java and Spring ecosystem for their invaluable tools and resources.

---

📬 Contact

For questions, feedback, or permissions, please reach out:

• Email: rohanuke1@example.com
• GitHub: @ARONAGENT
• Repository: Spring_AI

---

⭐ Show Your Support

🌟 Star this repo if you find it helpful! ⭐

Your support motivates continued development and helps others discover this resource.

---

Built with ❤️ by ARONAGENT

Making AI accessible, one Spring application at a time.