Government & Public Sector

A government system
crashing multiple times
a day.

Subdere ran on on-premise infrastructure with a system so outdated that compiling code required moving it from computer to computer. Crashes were daily and recovery was slow. We modernized the code, implemented CI/CD with GitLab, and migrated the architecture to High Availability — turning a fragile platform into one the IT team no longer needs to fight fires to keep alive.

Project Summary Completed case
0 crashes Stable platform with no daily interruptions
2 fronts Code and architecture modernized in parallel
High Avail. Fully independent QA and Production environments
CI/CD Automatic deployment with GitLab Runner — no manual steps
The Challenge

When the critical system
can't keep running like this.

In public infrastructure, every system crash isn't just a technical problem — it's a government process stopped cold.

🏛️

A Critical System at the End of Its Useful Life

Code with no path to update

Subdere ran its daily operations on a system built years ago, with an architecture and codebase so difficult to maintain that compiling required physically moving code from computer to computer. Crashes happened multiple times per day — with no fast recovery, no separate environments, no automation. In public infrastructure, every crash is a government process stopped cold.

🔧

No CI/CD, No Change Control

Manual and risky deployments

Without a centralized repository or deployment pipeline, every code change was a manual and risky operation. No rollback capability, no isolated test environments, and any error went straight to production.

🚒

IT Team in Permanent Firefighting Mode

Resources trapped in incident response

Instead of building improvements, Subdere's IT team spent its time restarting the platform after each crash. Modernization wasn't an option — it was an urgent necessity.

The Solution

Two fronts.
One result: stability.

We attacked the problem in parallel — first cleaning the code to make it sustainable, then redesigning the architecture so it would never again depend on a single machine.

01
Code

Source Code Cleanup and Modernization

We performed a complete analysis of the application, identifying and removing functions and dependencies that degraded performance and made the code unreadable. We implemented GitLab as a centralized repository and created a pipeline with GitLab Runner to automatically deploy changes to QA and Production environments. For the first time, the team had full control over the code lifecycle.

GitLab GitLab Runner Refactoring CI/CD
02
Architecture

High Availability with Separated Environments

We migrated the system to a High Availability architecture with completely independent QA and Production environments. We updated the JBoss and PostgreSQL versions, eliminating vulnerability risk and unlocking new capabilities in the stack. The infrastructure went from being the cause of the problem to being its solution.

High Availability JBoss PostgreSQL QA + Production
03
Automation

Automatic Deployment, Zero Manual Intervention

The implemented CI/CD pipeline allows code changes to deploy automatically and in a controlled manner whenever needed — no manual coordination, no human error risk. The IT team stopped being the bottleneck in the development cycle and returned to building improvements instead of just keeping the lights on.

Automation GitLab CI Continuous delivery Self-service
Results

A system the government
can rely on.

Project Results

Subsecretaría de Desarrollo Regional

Government of Chile · Public Sector

0crashes
stable platform after modernization — from multiple daily outages to none
Before

Daily Outages with No Control

On-premise system with code impossible to update. Multiple crashes per day, manual computer-to-computer compilation, and no fast recovery path. IT team in permanent firefighting mode.

What We Did

Code + Architecture + CI/CD

Source code cleanup and refactoring. GitLab as centralized repository with GitLab Runner pipeline. High Availability architecture with QA and Production environments. JBoss and PostgreSQL version upgrades.

After

Stable and Autonomous Platform

Zero operational crashes. Automatic deployments without manual intervention. IT team focused on improving the system instead of firefighting. Ability to update efficiently whenever needed.

Tech Stack

Technologies
deployed.

The stack implemented to modernize and stabilize Subdere's critical platform.

CI/CD
GitLab GitLab Runner
Centralized repository and automatic deployment pipeline
Application
JBoss
Application server updated to stable version
Database
PostgreSQL
Relational engine updated with high availability
Architecture
High Availability QA Environment Production Environment
Redundant platform with isolated environments

Is your platform crashing
more than it should?

We modernize critical systems — code, architecture, and automation — without stopping your operation. If your IT team spends more time fighting fires than building, let's talk.

Schedule a meeting →