Most teams think more context makes AI coding agents better.

In enterprise Java, it often does the opposite.

After working with Copilot, Claude, and lately IBM Bob + Bobshell, I learned one thing: your AGENTS.md is not documentation. It’s a control surface.

Keep it small. Move conventions to skills. Separate repo rules from policy.

Here’s the full breakdown:
https://www.the-main-thread.com/p/optimize-agents-md-ai-coding-java-enterprise

#Java #EnterpriseJava #AICoding #LLMs #SoftwareArchitecture

Most teams don’t start from scratch.

They have a working Quarkus app. Repositories. Services. REST endpoints. Production traffic.

In our latest episode, Alex Soto shows how to use IBM Bob to add LangChain4j to an existing Quarkus application — step by step, no rewrite.

It’s about controlled AI integration. Not hype.

Full article:
https://www.the-main-thread.com/p/from-standard-to-smart

#Java #Quarkus #LangChain4j #AIEngineering #EnterpriseJava #DevExperience

Hibernate with Panache Next just landed in Quarkus main.

One API.
Blocking + reactive unified.
Stateless sessions.
Type-safe @Find queries validated at build time.

This is a real shift in how we build persistence layers in Java.

I built a full bookstore REST API to show:
• Managed vs stateless sessions
• Reactive repositories
• What breaks under load

Full hands-on deep dive:
https://www.the-main-thread.com/p/quarkus-panache-next-jakarta-data-hibernate

#Quarkus #Java #Hibernate #JakartaData #EnterpriseJava

Software kann zum Beweismittel werden! #EnterpriseJava-Systeme stehen zunehmend im Fokus forensischer Analysen – Commit für Commit. Prof. Dr.-Ing. Stefan Wagenpfeil erklärt, wie man sich vorbereitet. Ist euer Entwicklungsprozess belastbar?

Jetzt lesen: https://javapro.io/de/code-vor-gericht-java-als-beweismittel/

Org charts are boring.
Until your company does another reorg.

This article shows how to build an interactive org chart with Java, Quarkus, PostgreSQL recursive CTEs, and D3.js — designed for orgs that change constantly (yes, even yours).

Hands-on. Full code. Real hierarchy queries.
https://www.the-main-thread.com/p/interactive-org-chart-java-quarkus-postgresql

#Java #Quarkus #PostgreSQL #SQL #WebDev #EnterpriseJava

Eclipse GlassFish 8 is Released
https://techlife.blog/posts/glassfish-8-release/

#JakartaEE #GlassFish #EnterpriseJava #OmniFish #Java21 #Java

LangChain4j has crossed an important line:
from “interesting library” → production Java AI infrastructure.

I wrote a long-form guide with 50 LangChain4j interview questions, covering:
• AI Services
• RAG & embeddings
• Tools & agents
• Memory & context limits
• Observability, cost, and security

Written from a production Java perspective — not Python demos.

https://www.the-main-thread.com/p/langchain4j-interview-questions-java

#Java #LangChain4j #AIinJava #EnterpriseJava #RAG #LLM #SoftwareArchitecture #DevExperience

Most persistence bugs are boring.

Renamed fields. Broken query strings. Invalid sort orders.
They compile. Tests pass. Production breaks later.

Jakarta Data in Quarkus 3.31 moves these failures to compile time.

This article explains why that matters, how Panache 2.0 fits in, and why the Hibernate annotation processor is the real story.

https://www.the-main-thread.com/p/jakarta-data-quarkus-compile-time-persistence

#Java #Quarkus #JakartaData #Hibernate #EnterpriseJava

Most toxicity filters fail because they treat language as text, not meaning.

In this article I show how to build semantic toxicity detection with Quarkus Guardrails using a dedicated classifier model, not regex or vague embeddings.

Multi-dimensional scoring. In-process ONNX. Production-grade Java.

👉 https://www.the-main-thread.com/p/semantic-toxicity-detection-quarkus-guardrails-java

#Java #Quarkus #AI #LLM #AISafety #EnterpriseJava