Loïc Magnette on JAVAPRO International

LangChain4j Agentic Workflows: From AI Calls to Multi-Agent Systems in Java

Wed, 08 Jul 2026 07:00:02 +0000

You have built AI features into your Java application. Your model is wrapped in a service, RAG is feeding it context, tools are wired, and calls are flowing. It works. Then requirements evolve. A single prompt-and-response is no longer enough. You need steps that follow each other, branches based on decisions, retries when things fail, and multiple actions running concurrently. The question shifts from “how do I call an LLM?” to “how do I orchestrate multiple LLM-driven tasks into a coherent system?”

A2A Building Interoperable AI Agents with Java

Wed, 18 Mar 2026 07:00:02 +0000

Why agents need friends?

We’ve recently seen a boom in AI “agents”. AI Agents are usually described as a service that talks to an AI model to perform some kind of goal-based operation using tools and context it assumes.

But most of these agents are still working in an isolated environment. You build an agent into your application and it offers some capabilities, but that’s it. You’re basically building a monolith with AI in it.

Bridging the Gap: Full-Stack Development Without the Headaches

Tue, 06 May 2025 07:00:03 +0000

Working on web projects often means dealing with two separate environments—one for the backend and another for the frontend—each with its own ecosystem. Developers frequently run into issues when dealing with the ecosystem of their counterpart, creating a needless wall between front-end and back-end developers. You have to use unfamiliar tools, follow API changes, create mocks, and have a hard time creating e2e tests. Overall, this makes development unnecessarily frustrating and a time-consuming process.

AI-Powered Form Wizards: Chat, Click, Done

Tue, 15 Apr 2025 10:00:23 +0000

Transforming Form Filling into a Conversational Experience

Forms are everywhere—tax declarations, job applications, or even signing up for a new service. Although some forms are simple, many include ambiguous fields, complicated logic, or subpar design. This may frustrate users and make them more likely to make mistakes. Completing paperwork shouldn’t be like solving a puzzle

Traditional forms, with their rigid structures and often confusing layouts, present a significant hurdle for users. Our objective was to dismantle this static paradigm and replace it with a dynamic, conversational interface. Instead of forcing users to navigate a pre-defined maze of fields, we envisioned an interactive experience where an AI assistant adapts in real-time. This approach fundamentally shifts the burden of data entry and validation. A user chatting with an AI can dramatically reduce errors and streamline the overall process. Imagine a conversation, not a questionnaire, where the AI guides you through each step.