0

Your cart

TOTAL excl.
TOTAL incl.
Pay
Go back

From packaged software to "Software Engineering": industrializing a market tool

REXAutomatisationCI/CD

2026-12-09 | 04:05 PM - 04:35 PM

Information

Tired of manually clicking through your software packages? Discover how our team applied Software Engineering practices to a proprietary market tool. GitLab CI/CD, Python code quality, Squish testing, and Splunk observability: a concrete case study to boost your time-to-market.

Configuring and setting up a commercial tool (software package) too often means manual intervention, a lack of traceability, no testing, and frustration for technical teams. Faced with this reality within our "Build Squad," we decided to take a radically different approach: treating the configuration of this proprietary product exactly as if it were standard application code, while injecting all the rigor of DevOps culture and Software Engineering.

In this case study, we'll give you a behind-the-scenes look at our software factory. We'll detail the steps, the architecture of our pipeline, and the tools we chose to tackle this challenge.

The presentation plan will be structured around three main themes:

  1. Code Quality Automation: Code reviews shouldn't be the neglected aspect of software packages. We'll explain how we automated our configuration reviews and project metric sharing using lightweight Python-based solutions, seamlessly published on GitLab Pages.
  2. Industrializing build and documentation: How we implemented a "Macro Build & Delivery" model with a Git workflow adapted to the constraints of a closed product. We'll see how creating shared libraries and the Doc-as-code approach with Docusaurus allowed us to minimize our development efforts while improving time-to-market.
  3. Validation and Observability: Our Power Trio: Once the code is industrialized, how do you ensure secure deployments? We'll present our automated testing stack combining Squish, Python, and Splunk to track down even the smallest regressions. Finally, we'll discuss the "bonus" of our approach: cross-analysis of instance usage via heatmaps in Splunk, allowing us to measure the actual adoption of our components.
This session is for anyone involved in the industrialization of third-party tools. You will leave with concrete methodological and technological keys to break down the silos of proprietary software.

Speakers