I didn’t try it with ollama. According to the Spring AI docs: https://docs.spring.io/spring-ai/reference/api/chat/functions/ollama-chat-functions.html#_how_it_works – it should work. Anyway, a functional calling feature is already deprecated in Spring AI… I’ll update this arrticle to describe replacement – Tool Calling

[404] Not Found – {“error”:”model \”mxbai-embed-large\” not found, try pulling it first”}

Thanks for careful reading. I changed it as you suggested

OpenAI doesn’t call your API over internet. Instead, it just suggests function calls, and your code executes them locally.

Your Spring Boot app sends a request to OpenAI with function definitions.
Then OpenAI responds with the function name and parameters.
Your app receives the response and runs the function locally. More details: https://docs.spring.io/spring-ai/reference/api/chat/functions/openai-chat-functions.html

Just a quick note: I found mistakes on the architecture section, you mixed the usage of StockService with WalletService both in the text and architecture image.

I should be written:
Chat model should call StockService funtion for each object in the list returned by WalletService function