All articles
Gather our latest posts and blog insights
36 posts

Client Extensions in Liferay: how the front end says goodbye to OSGi modules
What Liferay's Client Extensions are, their types, and how to build a custom element in React step by step, from the yaml to deployment.

Key rotation in complex architectures: order, steps and evidence
How to plan credential rotation without downtime: inventory, dependency order, dual credentials and evidence logging.

Headless Liferay with Next.js: OAuth2, Zod and caching in production
How we consume Liferay's headless APIs and Objects from Next.js: authentication, typing, revalidation and images.

Deploy a project to Vercel with Prisma and Postgres for free: step-by-step guide
How to set up and publish a side project with Next.js, Prisma and Postgres without paying anything, avoiding the typical serverless mistakes.

Liferay widget templates as code: from the workspace to production without touching the UI
Liferay widget templates live in the database: no Git, no review and copied by hand between environments. We explain how we manage them from the workspace with an OSGi module that, on deploy, creates or updates the ADTs and clears the template cache. With an appendix: the className of every widget template in DXP 2026.Q3.

Internationalized Next.js: how we set up i18n with next-intl
How we structured routes, messages, formatting and multilingual SEO with next-intl in alquila-sabanas, our Next.js project.

Caching in Next.js 16: use cache, revalidateTag and cacheLife without the drama
How the new caching model in Next.js 16 works, with use cache, cacheTag and cacheLife, and the most common mistakes when applying it.

Server Components vs React Query: do you need a client-side fetcher?
When a Server Component is enough to load data and when React Query pays off: mutations, optimistic UI, polling, and the hybrid pattern.

SSR, SSG, ISR, CSR and SWR: a practical guide to choosing well in Next.js
Four ways to render a page and a caching strategy that isn't the same thing. Differences, App Router examples, and when to use each one.

How we keep the AI that codes from seeing our passwords
A hook reviews every command before it runs in Claude Code so that no secret ever reaches the conversation with the AI.

AI Doesn't Reduce Technical Training, It Makes It Essential
Why generative artificial intelligence doesn't replace technical knowledge, but instead turns it into the differentiating factor in tech work.

AI agent security: why a prompt is not enough as a policy
A system prompt doesn't stop an attack. How to design real guardrails for AI agents in enterprise environments with n8n.

Agentic browsing: what we did so an AI agent knows how to use our website
More and more visits to a corporate website are not made by a person but by an assistant working on their behalf. That visitor gets the same thing a human gets, a page full of menus, scripts and text spread across components, and has to guess what is useful. Here is what we have published on www.julditec.com so that an agent can find out what we do, ask about it and get data, and why HTTP status codes were not enough to check it.

Evals and observability: how to know if your AI agent really works
Without evals or traces, an AI agent is a black box. We explain how to measure it, monitor it, and catch failures before the customer does.

Skills and Context Engineering: teaching an AI agent your way of working
How to document internal processes as reusable "skills" so AI agents work with your methodology, not with a generic one.

MCP (Model Context Protocol): how it connects AI agents with your data
What Model Context Protocol is, how its architecture works, and why it's changing the way AI connects with tools and real data.

What an AI agent really is: LLMs, tools, loops and agent harnesses
We break down the concept of the AI agent by explaining its four key pieces: the LLM, the tools, the reasoning loop, and the agent harness that holds it all together.

RAG: how to give an LLM its own knowledge without training it
Discover what RAG (Retrieval-Augmented Generation) is, how its architecture works, and how to implement it with embeddings, vector databases, and n8n without needing to train a model from scratch.

Jev and the Decision Layer: Do We Need an LLM for Every Step of an Agent?
Jev proposes separating thinking and generating from deciding in AI agents. We analyze its architecture, real-world cases, and why not everything needs a generative LLM.

How an LLM Runs: Inference, Quantization, KV Cache, and Context
Discover what technically happens when an LLM generates text: inference, model quantization, KV cache, and context window explained with practical examples.

Orchestrating AI agents in parallel with git worktrees: Orca's architecture
How Julditec coordinates multiple AI agents in parallel with git worktrees, durable messages, heartbeats, and an adversarial review round before merging anything.

How an LLM Really Works: From Tokens to Final Response
Discover the complete process behind language models: from tokenization to generating coherent responses. A technical yet accessible guide.

How to Transform a Traditional Portal into a Composable Platform with Liferay
Evolve your monolithic portal towards a composable architecture with APIs, Client Extensions and headless CMS. Practical guide to gain agility and autonomy.

Developing in the AI Era: Automate, Parallelize, and Document or Fall Behind
AI has transformed software development by multiplying productivity, but also pressure. Discover the three fundamental pillars to survive and thrive in this new context without falling into burnout.

Testing Strategy at JULDITEC: Guaranteeing Over 80% Coverage
Discover how at JULDITEC we guarantee maximum software quality through a comprehensive automated testing strategy, with over 80% coverage and continuous quality analysis.

Zustand in Liferay Client Extensions: Performance and Simplified Global State
Discover how to integrate Zustand in your React Client Extensions in Liferay to drastically improve performance, avoid unnecessary re-renders, and simplify global state management.

Devcontainers: The Foundation of Modern AI-Powered Development
Discover how devcontainers have become the standard for reproducible environments and why they are essential when working with AI in software development.

Adapt or Fall Behind: The Mindset Shift in the IT Sector
The IT sector is experiencing an unprecedented transformation. The real challenge is not technological, but mental: learning to constantly adapt in an ecosystem that changes faster than ever.

Liferay Headless: Modern Architecture for Decoupled Experiences
Discover how Liferay DXP has become a robust headless platform with REST APIs, OAuth2, and the new Language Service in 2026 Q1 for decoupled frontends.

Liferay MCP Server: Complete Configuration and Usage Guide with AI
Learn how to configure Liferay's MCP server to connect AI assistants like Claude with your platform's entire REST API. Manage content, users, workflows, and commerce with natural commands.

Intelligent Automation with n8n and AI: How We Transformed Document Management
Discover how n8n combined with artificial intelligence enables the automation of complex business processes. Real case: automatic invoice processing with AI classification, data extraction, and bank reconciliation.

How to Work with WebP Images in Liferay?
Steps to fully enable WebP support in Liferay, from dependencies to its integration in the media library and image selector.

Definition of Done in UI/UX Design: When a Task Is Truly Ready
A basic Definition of Done checklist for design, organized by visual quality, usability, accessibility, and validation.

Improving Liferay Search with Synonyms
Implementing synonyms in Elasticsearch and using them in Liferay to improve search result relevance step by step.

How to Organize a Design System
Structure by purpose, abstraction levels, and function to achieve efficiency, consistency, and a better internal experience.

What Is Liferay and What Is It For: 2026 DXP Guide
What Liferay is, what it's for, and how a CMS differs from a DXP. A clear guide with real use cases and how to implement it in your organization.