Java SDK + Spring AI integrations for #TypeSafe AI's #Jev. Not a chat model! Send it a state and typed questions, get numbers back with a confidence, in ~300 ms. Judge, self-refine + guardrail advisors, RAG triage, tool selection. Blog: spring.io/blog/2026/09... @spring-ai.bsky.social

Spring AI 2.0.1 is live 🎉 First maintenance release since 2.0 GA - 80+ issues/PRs, 7 CVEs, and some frequently-requested features landed: configurable tool call limits, OpenAI audio streaming, Google GenAI image generation, and a big round of streaming/tool-calling fixes. spring.io/blog/2026/08...

Honored to join the Agentic AI Foundation (AAIF) Ambassador cohort. AAIF is the home of AI projects, including the MCP. As one of the MCP Java SDK maintainers and Spring AI lead, I'm looking forward to seeing more developers bring AI into the Java and Spring ecosystem www.credly.com/badges/a2dfe...

Agentic AI Foundation Ambassador 2026 was issued by The Linux Foundation to Christian Tzolov.

This badge is earned by being selected as an Ambassador for the Agentic AI Foundation.

credly.com

LLMs: text-in, text-out. Enterprise code: structured-in, structured-out. #SpringAI 2.0 adds two switches on .entity(): • validateSchema() — retries with the error fed back • useProviderStructuredOutput() — pins the schema at the provider spring.io/blog/2026/06... #Java #AI #StructuredOutput

You could call tools. You couldn't build on top of tool calling! #SpringAI 2.0 lifts the tool loop out of the chat models and into the advisor chain - composable, observable, replaceable: spring.io/blog/2026/06... #Java #AI #MCP

🍃 From Assistants to Agents: Self-Improving Agentic Systems with Spring AI by @tzolov.bsky.social @ Spring I/O 2026 ▶️ Video: youtu.be/_ZZuzakjgxQ #springio26

From Assistants to Agents: Self-Improving Agentic Systems with Spring AI by Christian Tzolov

YouTube video by Spring I/O

youtu.be

What a discussion! It was a pleasure to moderate AI panel at @geecon.bsky.social Thank you @dylanbeatt.ie @tzolov.bsky.social @askowronski.bsky.social and @fmartin.ch hope that we manage to tackle a least a few important questions 😀

Excited to be speaking today at #GeeCON 2026! 🎉 📅 15:10 - Building Self-Correcting AI Agents AI agents are relentless context orchestrators - and the Advisor Pattern is all you need to compose them 🌱 #SpringAI #AI #Java 2026.geecon.org/schedule/

2nd talk at @springio.net today 🌱 — One abstraction rules them all 💍 The Advisor pattern is all you need for any Agentic Harness — Tool Calling, Memory, RAG, Guardrails, LLM-as-a-Judge & more. Self-Improving Agentic Systems with #SpringAI 📅 14:30–15:20 · Track 2 2026.springio.net/sessions/fro...

@tzolov.bsky.social introduces the @spring-ai.bsky.social roadmap at Spring I/O keynote! 🧵 #spring #java #springio26

File-based long-term memory for #SpringAI agents - no vector store, no database, just files. Plain Markdown files, MEMORY index, the agent manages it all via tool calls. Claude auto-memory compatible but portable to any LLM provider. 📖 spring.io/blog/2026/04... #Java #AI

Spring-AI deep-dive by @tzolov.bsky.social at #vdams26. Fast-paced with lots of info, at the same time slowly building up to show how all features are based on a small nr of core principles and techniques, very nice!

Prepping slides & demos for my Voxxed Days Amsterdam 2026 deep dive tomorrow Packing in Memory, Advisors, RAG, Tool Calling & MCP, Agentic Patterns, A2A & ACP, LLM-as-a-Judge, Guardrails… and somehow fitting it all into 2 hours! See you there 👉 amsterdam.voxxeddays.com/talk/?id=12905

Spring AI 2.0.0-M3 is now available and with it comes the ability to set metadata for tools and resources. And that makes MCP Apps possible. Read about how I’ve been tinkering with MCP Apps and how to blend AI chat with traditional UI design at spring.io/blog/2026/03...

Blending Chat with Rich UIs with Spring AI and MCP Apps

Level up your Java code and explore what Spring can do for you.

spring.io

🚀 Speaking at @amsterdam.voxxeddays.com — April 1–2! "Spring AI: From Fundamentals to Agentic Patterns" — a hands-on deep dive into context window curation, RAG, MCP, subagent orchestration, and LLM-as-a-Judge. All in Java. 🎟️ amsterdam.voxxeddays.com/talk/?id=12905 #SpringAI #Java #Agentic #AI

Java is #AI infrastructure front-runner ⚡ Benchmark of MCP server implementations across 4 major languages puts #MCP #Java & #SpringAI at the top of the performance charts: sub-ms latency, highest throughput, and the best CPU efficiency 📖 www.tmdevlab.com/mcp-server-p...

Subagent Orchestration delegate tasks to focused subagents with isolated contexts and run in parallel #SpringAI TaskTool provides a portable, model-agnostic implementation inspired by Claude Code's subagents 📖 spring.io/blog/2026/01... @spring-ai.bsky.social #AI #Java #Agents

🚀 Spring AI 2.0.0-M2 is here! ✨ Highlights: - Anthropic Custom Skills support, - Null-safe API with JSpecify, - New vector stores: S3, Infinispan, Bedrock KB, - Redis Semantic Cache, - Structured outputs for Mistral AI & Ollama, - MCP SDK 0.17.2 and bug fixes spring.io/blog/2026/01...

Spring AI 2.0.0-M2 Available Now

Level up your Java code and explore what Spring can do for you.

spring.io

Tomorrow, I’ll have the privilege to kick off the jChampions Conference, a free online event with sessions presented by Java Champions ☕️ I’ll talk about Modular RAG with Spring AI, Arconia, and Docling at 3:00 pm CET 🍃 www.youtube.com/watch?v=ZcB4... Schedule: jchampionsconf.com/schedule.html

Modular RAG Architectures with Java and Spring AI

YouTube video by JChampions Conference

youtube.com

#SpringAI is among the official adopters of the Agent Skills specification: agentskills.io It is nice to see a Java toolkit next to Claude, Cursor, Gemini, GitHub... Find more about the Spring AI Skills integration: www.linkedin.com/feed/update/... #Java #AI @spring-ai.bsky.social

If AI agent drops steps on complex tasks, add TodoWriteTool-a #SpringAI tool that enables LLMs to create, track, and update task lists during execution. It transforms implicit planning into explicit, trackable workflows. 📖Blog: spring.io/blog/2026/01... #Java #AI

#SpringAI AskUserQuestionTool - AI agents that ask clarifying questions before answering, gather requirements interactively, and create a specification aligned with your actual needs 📖Blog: spring.io/blog/2026/01... #AI #Java #Agent @spring-ai.bsky.social

Bringing Agent Skills to Java and #SpringAI 🚀 Modular, reusable capabilities for your AI agents -discovered at startup, loaded on demand. Portable across any LLM. Inspired by & compatible with Claude Code Skills. Part 1 of the Agentic Patterns series: spring.io/blog/2026/01... @spring-ai.bsky.social

🎉As 2025 comes to a close, I'm taking a moment to reflect on an extraordinary year of innovation with Spring AI - From MCP integration to agents, explainable AI, token optimization, ... Read the full story 👇 www.linkedin.com/pulse/2025-y...

🤖 Building AI agents? You can capture WHY the LLM makes decisions—not just what it decides. #SpringAI's Tool Argument Augmenter captures LLM reasoning alongside tool calls—useful for building memory-enhanced agents. No tool modifications needed! 📖 spring.io/blog/2025/12...

#SpringAI Tool Search Tool - Build smarter AI apps that discover tools when they need them 📖 Blog: spring.io/blog/2025/12... 🔗 GitHub: github.com/spring-ai-community/spring-ai-tool-search-tool ✅ On-demand tool discovery ✅ Multiple LLMs ✅ 34-64% token savings @spring-ai.bsky.social

JSON is the go-to format for LLM tool responses, but alternatives like TOON, XML, CSV, YAML can potentially improve token efficiency and performance. Find how to configure #SpringAI tool responses formats and optimize for your use case: spring.io/blog/2025/11... @spring-ai.bsky.social

A few thoughts from Chicago.

🚀 Spring AI 1.1.0 GA Released 🔧 Better MCP Integration ... 🔐MCP OAuth2 server protection 🔄 Recursive Advisors ...📦 Developer Experience: ... Over 850+ commits Kudos to the entire Spring AI community! 🙌 Find more: spring.io/blog/2025/11... @spring-ai.bsky.social

🤖⚖️ LLM-as-a-Judge with #SpringAI Evaluating LLM output is challenging. Traditional metrics fall short, and human evaluation doesn't scale. LLM-as-a-Judge uses LLMs to evaluate AI-generated content, matching human judgment 📖 spring.io/blog/2025/11... 🛠️ github.com/spring-proje...