From legacy code to modern standards: transforming Holafly’s infrastructure
Client
Holafly
services
Digital Transformation
industry
Telecommunications
The Challenge
Holafly’s commercial success came so quickly that its technology couldn’t keep pace. The company faced critical technical debt that threatened its ability to continue innovating and growing. In this context, they needed a technology partner to help them:
—Eliminate concentrated knowledge dependency: Only a few developers understood the critical system operations, creating bottlenecks and operational risk with any absence.
—Implement quality standards: With 0% test coverage and a complete absence of typing, every production deployment was a leap into the void, multiplying errors and compromising stability.
—Modernize observability: Deprecated protocols and a lack of traceability prevented incident identification and resolution, forcing the team to operate practically blind.
—Unblock critical projects: The new support platform had been paralyzed for months, never making it to production, generating frustration and eroding team confidence.
The Solution
Orbitant chose a pragmatic approach: implement testing, typing, and observability without rewriting existing code or stopping ongoing development. The strategy focused on three fundamental pillars:
—Incremental typing and integration tests: Implementation of typing through jsconfig.json (respecting the team’s reluctance toward TypeScript) and creation of black-box integration tests that elevated coverage from 0% to 90% in critical services, with automated pipelines and analysis through SonarCloud.
—Reverse engineering of the proprietary library to access source code, adoption of the Open Telemetry standard, and optimization of trace architecture, eliminating unnecessary use of PM2 in Kubernetes.
—Complete restructuring of the blocked project: Production release of the support platform in less than 1 week (after months blocked), with complete API restructuring, creation of 6 screens and 8 endpoints in 2 weeks, strategically using AI to accelerate without compromising architecture.
Every decision responded to the same principle: deliver immediate value while laying the foundation for sustainable growth.
Tech Stack
Nodejs, React & Python
GITHUB ACTION / CI
OPEN TELEMETRY
The Results
These changes unlocked capabilities the team simply didn’t have before. They went from flying blind to having full system visibility, from projects stalled for months to shipping in days, and from knowledge siloed in a few people to a mature setup where any developer can jump in and contribute with confidence.
Business Impact
90% test coverage
The system went from having no automated tests to achieving between 50% and 90% coverage in critical services. Integration tests not only validate the code but document the expected behavior of each endpoint. Automated pipelines prevent unvalidated code from reaching production, detecting typing errors, misused variables, and methods with incorrect parameters before deployment.
delivery velocity
A project blocked for months with mocked data and non-functional APIs went to production in less than 1 week. All APIs were restructured, 6 new screens and 8 functional endpoints were created with their integration tests. In 2 weeks, a complete MVP was delivered that allowed the support team to operate effectively.
Modern traceability
Migration from deprecated protocols to the Open Telemetry standard transformed observability capabilities. Unnecessary use of PM2 in Kubernetes that caused trace loss was eliminated. The team went from operating blind, without knowing what was happening in production, to being able to follow each request through all microservices and identify the exact origin of incidents.
Knowledge transfer
Through pair programming, best practices training, and knowledge spaces like IT.all, Holafly’s team gained autonomy to create and maintain tests, implement correct typing, and follow established standards. Knowledge stopped being isolated in a few developers to being documented in code and shared across the entire technical team.
The Impact in Numbers
+90%
test coverage in critical services
×8
Faster production releases
+4
services with new tracing systems
