Generative AI applied to software quality for fiber network management
Executive Summary
Geosite Telecom is a platform created by Geosite Tecnologia based on geolocation and used by over 500 internet providers to manage all documentation and maintenance of fiber optic networks (FTTH). It is a highly complex product that supports the daily connectivity operations of thousands of end-users.
Due to the product's long history, the system had accumulated complexity, making the testing routine and updates a challenge whenever new features were released.
With the Genius platform, Squadra structured a quality and test automation approach based on Generative AI to support everything from the definition of the technical strategy to the implementation of automation, ensuring the product's scalability, reliability, and continuous evolution, and reducing the risk of critical failures that could impact customer network management.
Furthermore, the case accelerated the team's learning curve, establishing a scalable architecture for a new automation agent within the Genius platform, which can be reused in projects for various clients.
The Challenge
The solution is essential for the operational continuity of over 500 internet providers and telecommunication companies, being vital for emergency repairs of network outages and infrastructure expansion. Although developed over more than twenty years, the solution keeps pace with market demands and undergoes continuous evolution, requiring agility and security in updates. To ensure that new features are implemented without compromising the platform's stability, test automation has become a strategic necessity to guarantee quality, reduce risks, and sustain the solution's continuous growth.
Main challenges
Pressure for constant technological evolution without loss of stability.
Need to continuously evolve the product without causing regressions in existing critical functionalities.
Need to structure a scalable automation architecture from the outset.
Ensuring quality in deliveries without increasing manual validation effort.
High risk of side effects in system updates due to code complexity.
The Solution
Genius in action
Squadra adopted a strategic approach based on the Genius platform, using Generative AI as a co-pilot throughout the solution's development.
The work began with understanding the business context and the platform's technical characteristics. Based on this analysis, the AI supported the evaluation of the existing architecture, recommended the most suitable strategy for automation, and guided the definition of the framework best aligned with the system's needs.
With this guidance, a complete test automation architecture was structured using best engineering practices, including organization based on Page Objects, separation of responsibilities, prioritization of critical network flows, use of fixtures, integration with support APIs, and automatic generation of evidence and reports.
Throughout the project, AI was also used as a learning tool, supporting the team in understanding technical decisions, applying best practices, and incrementally resolving challenges encountered.
This approach allowed for accelerated automation implementation while maintaining the product's operational integrity regardless of updates.
Main deliverables
Complete structuring of the automation architecture: the AI contributed to a technical diagnosis of the platform, the definition of the automation strategy, and the selection of the framework best suited to the system's context.
Intelligent and agile automation: the main critical application flows were automated via AI.
Living documentation: the interaction with AI itself served to map and document business rules that were previously only tacit knowledge within the team.
Implementation of traceability mechanisms, evidence, and automated reports.
Creation of the foundation for the Test Automation Turing Bot for reuse in new business contexts.
Results
The implementation of the automation strategy reduced the time to structure new test scenarios from weeks to just four days. This allows Geosite to deliver new features to the market much more quickly, without compromising the time-to-market for providers.
Quality of the Geosite Telecom product: with a more rigorous and predictable validation process, we have increased the platform's quality, accelerating the evolution cycle of a critical platform for an entire telecom ecosystem and ensuring the operational continuity of thousands of field teams and millions of connections managed by over 500 internet providers.
Reliability and security barrier: the scalable testing environment now acts as an additional layer of security, identifying failures before they impact provider operations and allowing new features to be delivered with greater stability, reducing the risk of regressions and downtime during updates.
Business Agility: the product is now even better prepared to evolve continuously at the speed required by the telecom market, preserving the robustness needed to serve hundreds of internet providers who depend on the platform for their daily operations.
A reusable AI asset: the initiative has become a reusable asset of the Genius offering, an intelligent Test Automation agent (turing bot), prepared to accelerate future automation initiatives in different business contexts.
Want results like these?
Talk to our experts and find out how AI can transform your business.
