Tutorials and deep-dive implementations covering artificial intelligence and machine learning concepts in Java and other languages.
Build an MCP server with only the official Java SDK 2.1.0: tools and resources, a fat jar for Claude Desktop over stdio, Streamable HTTP on Jetty, and a Spring AI client connecting to it. Every output comes from a committed transcript.
Anthropic's five agentic workflows (prompt chaining, routing, parallelization, orchestrator-workers, evaluator-optimizer) as plain Spring AI ChatClient code, tested offline with a scripted ChatModel that records every prompt, with the failure modes and the call-count cost of each.
A Spring AI streaming chat UI with no React: ChatClient to a Flux of server-sent events, an htmx front end, per-session memory, a Stop button that really cancels the model call, and back-pressure, all measured against a stub OpenAI-compatible server (no real provider was run).
Point one Spring AI 2.0.2 OpenAiChatModel at Docker Model Runner and at Ollama by changing only base-url and the model name. Same GGUF model, same prompts: decode speed, time to first token, container memory and tool-call success, with transcripts. Measured on a CPU-only shared sandbox with the real docker/model-runner image; GPU and Docker Desktop paths were not measured and the post says so.
A Spring Batch 6 chunk step calls a Spring AI ChatClient for 100,000 generated support tickets. Resilience4j rate limiting and retry, skip-and-record for malformed JSON, a restart after kill -9 in a second JVM that ends byte-identical, and a per-chunk cost ledger, all run against a deterministic simulated model with every number taken from a committed transcript.
A green build does not mean the agent did the right task. A runnable Spring Boot 4.1.1 example shows an agent-style change passing its own tests while failing 3 of 4 acceptance tests, followed by the instruction file, skills, test-first prompts, review checklist and delegation limits that prevent it, with a reusable AGENTS.md template.
A measured look at agent-style first-draft Selenium tests in Java: 120 headless-Chromium runs show the draft passing 20 of 30 on the page it was written for and 0 of 30 after a redesign, and a repaired suite passing 60 of 60. Honest about what was and was not an agent.
A pull-request review bot in Spring Boot 4.1.1 and Spring AI 2.0.1: verify the webhook signature, chunk the diff with new-file line numbers, review each chunk, drop comments on lines the model invented, and keep the bill down with a cache and a chunk cap. Runs with a scripted model and no API key.
Build two agents that hand work to each other with Google's Agent Development Kit for Java 1.11.0: an LlmAgent with a tool, a SequentialAgent, and session state as the hand-off. Runs without an API key, using a scripted stand-in model that is labelled as such.
Draw an agent as a graph of nodes, edges and shared state, pause it for a human with a checkpoint, and resume it later. A loan-approval agent on LangGraph4j 1.9.5, with the one default (releaseThread) that silently erases a finished run.