Spring AI Training Course

1. Introduction to Spring AI

• Creating and configuring projects
• The role of prompts and prompt submission
• Writing initial tests
• Selecting appropriate models
• Configuring models
• Overview of Spring AI capabilities

2. Understanding responses

• Verifying the relevance of answers
• Assessing runtime accuracy

3. Prompt details

• Utilizing prompt templates
• Defining custom prompt templates
• Comprehending context
• The significance of the role parameter
• Influencing response generation through options
• Streaming and formatting output
• Interpreting metadata in responses

4. Leveraging your data and documents

• Understanding RAG (Retrieval-Augmented Generation)
• Setting up vector stores and loading documents
• Implementing a basic RAG workflow
• Implementing RAG using advisors
• Utilizing modular RAG capabilities

5. The role of memory in AI

• The necessity of memory
• Adding and configuring memory for conversation support
• Managing conversation IDs
• Implementing persistent memory
• Storing chat memory in vector stores

6. AI Tools

• Enabling tools in applications
• Understanding tool capabilities
• Developing and deploying tools
• Using functions as tools

7. The Model Context Protocol (MCP)

• The need for MCP
• Working with MCP Clients
• Developing MCP Servers
• Integrating databases and tools for MCP Servers
• Understanding HTTP and SSE (Server-Sent Events) transport
• Exposing prompts and resources

8. Monitoring operations

• Enabling actuator metrics
• Monitoring vector store operations
• Tracking model interactions
• Counting tokens
• Integrating with Prometheus and creating dashboards
• Tracing AI operations

9. Safeguarding generative AI

• Controlling document access via RAG
• Securing tools
• Mitigating adversarial prompting
• Moderating user input

10. Common generative patterns

• Content summarization
• Message translation
• Sentiment analysis

11. The role of Agents

• Defining an agent
• Implementing agentic workflows
• Chaining prompts, task routing, and parallelization
• Agent access via MCP