Observability through a telemetry framework for distributed web apps: error and performance monitoring for Technogym Mywellness CRM

Modern distributed systems and micro-services architectures present unprecedented challenges for enterprises in understanding application behavior in production. Traditional monitoring approaches, that rely on known failure modes, are insufficient to face the unpredictability of contemporary digital ecosystems. Basic monitoring can intercept the existence of the issue, but not rebuilding the reasons behind it, which has to be investigated manually. This is the reason why observability has emerged to support digitalized companies in this situation, since it is capable of inferring internal system states from external outputs such as logs, metrics and traces, without any prior knowledge of specific failure scenarios. This thesis explores the practical implementation of observability frameworks in entrepreneurial contexts, addressing the fundamental shift from reactive troubleshooting to proactive system intelligence evaluating modern observability platforms. Technogym Spa serves as case study for this research. Specifically, the core objective was providing users with high quality digital products and assistance. Before this infrastructure, the customer experiencing a problem while using the CRM services had to report it to the technical assistance. Even though developers were fast in solving the errors respecting the one-day deadline for high criticality, the whole process to make them aware of the alert expanded the time window during which the client could not access the broken parts of the application, lasting even days. They wanted to avoid it to add more value to the customer. The implementation focuses on the Mywellness CRM Progressive Web Application, demonstrating how observability principles can be applied to React-based frontend applications to enable proactive error detection and performance optimization while respecting strict privacy regulations.