A Systematic Literature Review: Application Programming Interface (API)in the Context of Modern Software and Integration with Artificial Intelligence

Abstract

This study aims to systematically review and analyze the development, roles, and applications of Application Programming Interfaces (APIs) within the context of modern software, particularly their integration with artificial intelligence (AI), blockchain technology, and low-code platforms to enhance system performance, security, and rapid development capabilities. The research was conducted using the Systematic Literature Review (SLR) methodology, following the PRISMA framework and employing the PICO model to define the research scope. Literature was retrieved from major academic databases including PubMed, IEEE Xplore, ACM Digital Library, Scopus, and Google Scholar, yielding a total of 450 articles. After applying inclusion and exclusion criteria and assessing the quality using the Critical Appraisal Skills Programme (CASP), 18 high-quality articles were selected for detailed analysis. The results revealed that AI-driven APIs demonstrated 35% higher performance in data analysis tasks compared to traditional APIs, although they exhibited an average latency of 200 milliseconds. Blockchain APIs enhanced system security by up to 80%, but at the cost of a 60% reduction in throughput. Meanwhile, low-code APIs reduced development time by 50%, though they supported only 20% of complex tasks. These findings highlight both the potential and the limitations of various API approaches and suggest that future development should focus on reducing latency, increasing throughput, and improving flexibility to meet the demands of modern software systems effectively.