Building And Monetizing Generative AI Plugins With Spring AI (SpringOne 2024)

The behavior of OpenAI GPT models can be enhanced and augmented with plugins. For instance, the Expedia plugin equips the AI with up-to-date information on lodging prices and availability, letting you book your next trip directly through the ChatGPT interface. In this hands-on session, you will learn how to create custom Spring AI plugins and launch them on the OpenAI marketplace. We'll start with the basics, developing a RAG-based generative AI application that uses an LLM, vector database, and Spring AI to search for information requested by users. Then, we'll introduce capabilities that will let the AI perform actions following user prompts. Finally, we'll turn the Spring AI app into a plugin and discuss how to monetize it on the OpenAI marketplace.

Speaker: Denis Magda, Director, Developer Relations, Yugabyte

[kappaj]

Couple of questions:
Why would you store the vector data in the same row as the rest of the data?
Also, is Spring JPA not a much better abstration to use for getting the data?
Would be nice have a look at the source code for the insert/update/delete stuff as well.
Most of the demo's uses the ChatClientBuilder and not the ChatModel - is there a specific reason for this?

[Spring-ng]

excellent!!! is the source code available?