From AI agents shipping features overnight to JVM performance breakthroughs and evolving Java tooling.
🎯 Featured Article
🤖 Translating a Website into 8 Languages with AI Agents in One Night
A practical experiment in AI-driven development: using Claude Sonnet and fleets of GitHub Copilot agents to internationalize an entire Java website — from specification to deployment.
⬢ Showcases agent collaboration — how multiple coding agents can coordinate tasks like extraction, translation and deployment.
⬢ Highlights AI-assisted internationalization — how AI can automate complex engineering workflows traditionally requiring weeks of manual work.
⏩ TL;DR (Quick Recap)
⬢ Adopt AI-assisted workflows as agents increasingly handle large engineering tasks such as code generation, translation and security discovery.
⬢ Leverage JVM innovations like the Vector API and new Java 26 capabilities to unlock performance improvements.
⬢ Revisit architecture choices with evolving frameworks, migration tools and infrastructure patterns.
⬢ Recognize that engineering impact often comes from simplicity, performance and maintainability — not complexity.
☕ JVM Corner
⚡ Optimizing Recommendation Systems with JDK’s Vector API
Netflix improved recommendation scoring performance using the Java Vector API.
⬢ Apply vectorization techniques to batch calculations and reduce CPU usage in data-heavy services.
⬢ Optimize memory layout to unlock hardware acceleration benefits in modern JVM workloads.
🧰 “j2k-vscode”: New Java to Kotlin migration tool from JetBrains for VS Code
A new LLM-powered converter helps migrate Java codebases to Kotlin directly from VS Code.
⬢ Use LLM-based conversion to automate large-scale Java-to-Kotlin refactoring.
📦 Hardwood: A New Parser for Apache Parquet
A lightweight, high-performance Java parser for Apache Parquet files targeting Java 21+.
⬢ Improve data processing performance when working with Parquet in JVM pipelines.
🌐 HTTP Client Updates in Java 26
Java 26 introduces updates to the modern java.net.http.HttpClient.
⬢ Enhance networking capabilities with improvements to the standard HTTP client.
🚀 Java 26: What’s New?
A walkthrough of the upcoming Java 26 release and its 10 JEPs.
⬢ Review incremental platform improvements across language and runtime features.
🧠 Data Oriented Programming, Beyond Records
An exploration of how Java may evolve data-oriented programming beyond records.
⬢ Explore potential language evolution from the OpenJDK Amber discussions.
🍃 Spring Updates
🤖 Overview of MCP Annotations in Spring AI
Spring AI introduces annotations that expose Java methods as AI tools and prompts.
⬢ Simplify AI integration with declarative configuration instead of manual schema setup.
🔄 Things I miss about Spring Boot after switching to Go
A developer reflection on the productivity advantages of the Spring ecosystem.
⬢ Compare trade-offs between Java frameworks and Go simplicity.
🗄 Flyway vs Liquibase with Spring Boot: Which One To Choose?
A practical comparison of two popular database migration tools.
⬢ Evaluate migration strategies based on project complexity and governance needs.
🏗 Adapter Pattern in Spring Boot: A Hexagonal Architecture Perspective
A guide to implementing the Adapter pattern within hexagonal architectures.
⬢ Decouple application logic from infrastructure and external services.
⬢ Apply architectural boundaries that improve maintainability and testability.
🔍 Extra Reads
AI & Developer Tooling
🤖 Code Mode: give agents an entire API in 1,000 tokens — Introduces a method for compressing large APIs into compact contexts usable by AI agents.
🧠 Introducing GPT-5.4 — Presents a new frontier model optimized for complex professional tasks.
🧑💻 Cursor Joined the ACP Registry and Is Now Live in Your JetBrains IDE — Brings an AI coding agent directly into JetBrains IDEs through the Agent Client Protocol.
⚙️ Go is the Best Language for AI Agents — Argues that Go’s model makes it a strong candidate for building AI agent systems.
Architecture & Infrastructure
🔎 How we rebuilt the search architecture for high availability in GitHub Enterprise Server — Describes the redesign of GitHub’s enterprise search infrastructure for resilience and durability.
☁️ Ready-to-use virtual clusters — Explains how virtual Kubernetes clusters enable isolation without the overhead of separate clusters.
🎬 FFmpeg at Meta: Media Processing at Scale — Shows how Meta uses FFmpeg to power large-scale video processing workflows.
Security
🔐 Password managers less secure than promised — Research reveals vulnerabilities in several cloud-based password managers.
🛡 Partnering with Mozilla to improve Firefox’s security — Demonstrates how AI models helped identify multiple high-severity vulnerabilities in Firefox.
Career & Engineering Thinking
🧠 Nobody Gets Promoted for Simplicity — Examines how engineering incentives often reward complexity over elegant solutions.
📈 The Habits That Separate Elite Programmers From Forever-Mediocre Developers — Highlights the learning and discipline patterns that drive long-term developer growth.
🔑 What is OAuth? — Breaks down OAuth in approachable terms and clarifies its core authentication and authorization concepts.
Originally posted on marconak-matej.medium.com.