Slides & recording from my @devoxx.com deep dive: "Agent Loops, Decoded: Fundamentals to Agentic Patterns in Spring AI"

🎥 Recording: youtu.be/Q91obeWeWNQ

📊 Slides: speakerdeck.com/tzolov/agent...

#SpringAI #Devoxx #AI #Java #Agentic

📚 Zeit für neuen Lesestoff!

☕ Was bringt #Java27?
🤝 Was steckt hinter W-JAX Together?
🤖 Wie funktioniert #SpringAI 2.0?

Entdecke unsere 3 Leseempfehlungen im JAX-Blog! 👇

🔗 https://jax.de/blog/?utm_source=bluesky&utm_medium=social&utm_campaign=content

#jaxcon

LLM-as-a-Judge is a powerful evaluation technique, but what if you don’t need another LLM to make the call? With a System One model, you can fact-check AI responses using focused, typed judgments with less latency and lower cost.

medium.com/@thetalkinga...

#SpringAI #Java #SystemOne #AIEvaluation

SPEAKER DETECTED: Dan Vega @danvega.dev

Bring your laptop and build with Spring AI - from model abstraction and structured output to context engineering and tool calling.

Hands-on lab: “Spring AI in Practice”

www.jfokus.se #SpringAI #Jfokus

Your users don't want a workflow canvas. They want "if the subject says invoice, post to #finance."
Part 7 of @bytechef.bsky.social Embedded: build workflows in your own UI. One POST to create, one PUT per node to bind connections. ByteChef runs the rest.

blog.bytechef.io/blogs/embedd...

#springai

🚀 Hi, Spring fans! I’m writing from legendary Devoxx Belgium in Antwerp. Plus: Spring AI modular RAG, TypeSafe Jev, OneLens, podcasts, best practices, and more! Let’s go! #SpringBoot #SpringAI #DevoxxBelgium

https://spring.io/blog/2026/10/06/this-week-in-spring-october-6th-2026

Your AI agent tries to search Gmail. The user never connected it.

Instead of an error, the MCP server returns a connection link. The user connects, and the agent retries.

@bytechef.bsky.social Embedded part 6: MCP Chat.

#MCP #SpringAI
blog.bytechef.io/blogs/embedd...

Run System One models locally with Spring AI! In this recipe, see how Nimble and Laya deliver fast, focused AI judgments without a general-purpose LLM.

medium.com/@thetalkinga...

#SpringAI #Java #SystemOne #LocalAI #Ollama #Nimble #Laya

🤖 Spring AI 2.0

Thorben Janssen zeigt, wie ChatClient, Prompting, Chat Memory & Structured Output die Entwicklung produktiver LLM-Anwendungen mit Spring vereinfachen.

👉 https://f.mtr.cool/3th77ztrq2  

#SpringAI #Java #AI

🤖 What if your if statement could understand context?

🔎 Spring AI collects evidence.
🧠 Jev evaluates it.
☕ Java decides.

🎥 www.youtube.com/watch?v=-PQj...

#Java #SpringAI #Jev

Spring AI Modular RAG + TypeSafe AI Jev 🧩

LLMs rewrite and expand the query. Vector search finds 6 chunks. #Jev keeps the 2 that actually answer it.

📝 Blog: spring.io/blog/2026/10...
🎤 At @devoxx.com next week? Join my deep dive: m.devoxx.com/events/dvbe2...

#RAG #Java #SpringAI

220,000 tokens and 7 LLM calls down to 5,579 with a JSONata workflow, in a fraction of the time. Interpretable workflows beat code mode, no sandboxes needed. #JSONata #Tokens #SpringAI

Everyone says code mode solves MCP round trips. This dev disagrees: sandboxing, security, and LLM-written executable code on the fly? Here's the better path. #MCP #AIAgents #SpringAI

Every agentic app hands the LLM the raw tool response, even when it only needs a fraction. The fix: filter tool output before it hits the model. #SpringAI #AIAgents #Tokens

What if you don’t need an LLM to generate an answer, but simply need AI to make a judgment? In my latest Spring AI Recipe, I explore Jev, System One models, and fast, structured AI judgments.

medium.com/@thetalkinga...

#SpringAI #Java #Jev #SystemOne

Tool-search tool can raise your bill if it's misused.

Same task costs 68% more.

It pays off only when you aggregate multiple MCP servers and tool definitions alone eat thousands of tokens per request. With few tools, or when you defer ones you constantly use, it's overhead.

#LLM #Java #SpringAI

Why should a Java-based STDIO MCP server need a build step before a client can run it? With JBang, it doesn’t. You can build Spring AI MCP servers and clients that run straight from Java source.

medium.com/@thetalkinga...

#SpringAI #MCP #JBang #Java #SpringBoot

Indexing every tool behind your tool search tool is a bad practice I see a lot. Keep your go-to tools non-deferred.

#LLM #SpringAI #Java

Writing a book about Spring Boot 4 reinvigorated his love of coding. He's shipping a new Spring AI Agents MD starter to the community soon. #SpringBoot #SpringAI #Java

Contextual Retrieval RAG: cómo Anthropic arregla el chunking

¿Tus chunks de RAG pierden contexto? Contextual retrieval RAG con Spring AI y Virtual Threads baja fallas de recuperación 49% sin fundir el presupuesto

#contextualretrieval #rag #anthropicclaude #springai #virtualthreadsjava