Building AI-Driven Spring Applications With Spring AI

Imagine an AI that checks your smart fridge for ingredients before generating a recipe. See how to build it using Spring AI, function calling, and a vector database.

#1about 3 minutes

The evolution from machine learning to generative AI

Generative AI builds upon machine learning and deep learning by creating flexible, reusable models applicable to any domain or industry task.

#2about 3 minutes

Understanding foundation models, prompts, and tokens

Large language models (LLMs) use prompts and tokens to generate text, images, and other content based on complex probability calculations.

#3about 3 minutes

Introducing Spring AI for generative AI applications

Spring AI simplifies building generative AI applications by providing abstractions for models, vector databases, and advanced patterns like RAG.

#4about 6 minutes

Building a basic recipe finder with Spring AI

Use a ChatClient and prompt templates in Spring AI to easily call a large language model and map its JSON output to a Java object.

#5about 4 minutes

Switching AI models with only configuration changes

Spring AI's abstraction layer allows you to switch between different large language models, like from Ollama to Azure OpenAI, by only updating dependencies and application properties.

#6about 7 minutes

Enhancing prompts with real-time data using function calling

Implement function calling in Spring AI by defining a function as a bean, allowing the LLM to invoke your application's code to retrieve up-to-date information.

#7about 5 minutes

Implementing retrieval augmented generation with a vector store

Use Spring AI's vector store APIs and a QuestionAnswerAdvisor to implement Retrieval Augmented Generation (RAG), enriching the LLM's context with your own business data.