Diseño e Implementación de un Agente Inteligente de Pruebas Funcionales en APIs REST para Agilizar la Automatización en Calidad de Software

Abstract

This undergraduate thesis proposes a technological solution to the imbalance between the speed of agile development and the capacity to execute exhaustive testing. The central problem lies in the high burden of repetitive manual tasks and the high maintenance cost of conventional automation scripts, factors that limit the efficiency of Software Quality Assurance (SQA) teams. The fundamental objective of the research is to design and build an intelligent agent capable of assisting in the planning, design, execution, and reporting of functional tests for REST APIs. To achieve this, the system named Omega Testing was implemented, using a modular architecture composed of a React frontend and a NestJS backend with SQLite data persistence. The innovative core of the system is the integration of an Artificial Intelligence (AI) service based on OpenAI models, which allows natural language requirements to be transformed into automated test artifacts under the BDD (Behavior-Driven Development) approach and Gherkin language. The implementation was carried out using an applied research methodology and an agile framework structured into 12 sprints. The system centralizes modules for project management, endpoint inventory, test suite orchestration, and a monitoring dashboard with real-time events via Server-Sent Events (SSE). As a result, the evaluation with industry experts confirmed that 70% of the participants perceived a time reduction of between 31% and 70% in the generation and execution of tests. In conclusion, the sources demonstrate that the use of intelligent agents optimizes work times, reduces manual effort, and improves the coverage and precision of validations in continuous delivery environments.