The Main Thread

The Main Thread

Home
Podcast
Notes
Java
Java AI
Quarkus
FaQ
Archive
About
Why AI Coding Agents Still Need Clear Specs
Loose prompts look cheap until review, rework, token cost, and multi-agent drift show up; the useful middle is validated acceptance criteria, executable…
19 hrs ago • Markus Eisele
Quarkus Cache Invalidation Rules You Need Before Production
Build a small pricing service that shows when Quarkus caches resolved reactive values, when `@CacheInvalidateAll` skips invalidation, and when to force…
Apr 26 • Markus Eisele
Create Your First Quarkus SBOM with CycloneDX
Build a small Quarkus service, inspect distribution and dependency SBOMs, validate CycloneDX output, and archive the evidence in CI.
Apr 25 • Markus Eisele
Build Your First Real Java RAG Pipeline with Quarkus and Docling
Turn messy enterprise documents into structured retrieval with Docling, pgvector, Ollama, readiness checks, and guardrails in one local Quarkus…
Apr 24 • Markus Eisele
How to Run IBM Bob on a Remote Linux Machine from Your Mac
A step-by-step guide to using IBM Bob over SSH with a Podman-based remote host, so you can work closer to sensitive environments without turning your…
Apr 23 • Markus Eisele
Stop Letting AI Guess in Your Java Repository
Learn how Java developers can ground AI coding tools with JDTLS, MCP, and repository conventions to reduce bad code guesses.
Apr 22 • Markus Eisele
Build a Digital Credentialing Platform with Quarkus
Most badge systems look simple at first.
Apr 21 • Markus Eisele
The Hidden Cost of AI Coding for Senior Java Developers
Why AI-generated code feels fast, but shifts the real work into review, judgment, and mental overload in enterprise Java teams
Apr 20 • Markus Eisele
How to Review Agent System Prompts Like Production Infrastructure
A practical framework for grading coding-agent system prompts on grounding, continuity, safety, decomposition, and efficiency before they break real…
Apr 19 • Markus Eisele
AI Made Coding Faster. History Says That’s When the Real Problems Begin.
From Toyota’s production line to induced demand, the lesson is the same: the bottleneck always moves
Apr 18 • Markus Eisele
Chatbots Talk. Real AI Agents Schedule Work.
I met Ronald Dehuysser at Jfokus in February. We talked about Java, background processing, and the kind of problems that look simple until you have to…
Apr 17 • Markus Eisele and Ronald Dehuysser
Build a Streaming AI Chat in Java with Quarkus, Vaadin, and LangChain4j
A hands-on guest post by Sebastian Kühnau showing how to stream LLM responses token by token in a pure Java UI with Vaadin Flow and Quarkus.
Apr 16 • Markus Eisele
© 2026 Markus Eisele · Privacy ∙ Terms ∙ Collection notice
Start your SubstackGet the app
Substack is the home for great culture