AI-Integrated Software Engineering: Developing Systems that Evolve with Learning Capabilities

Abstract

The study examines AI-native systems that can be developed with the help of a Random Forest classifier to replicate the method of intelligent decision-making. The data contained variables like "user-behavior" and "performance-metric." The model obtained an accuracy of 97, and the factor analysis of feature importance shows that both user behavior and system performance are important factors that influence the outcome of decisions. The confusion matrix indicated that the model performed well with little misclassification. Findings highlight the prospects of AI-native systems in practice, but the adaptability and performance of systems in real-world situations need additional research using real-world data and continuous learning solutions to promote tighter integration of systems.