A Conceptual Framework for the Evolution of Application Development: From Manual Coding to Agentic Development

Purpose: This study proposes a conceptual framework to explain the evolution of application development from traditional manual coding to the emerging era of agentic development. While previous studies have primarily examined manual programming, low-code development, AI-assisted coding, or agentic software engineering as separate topics, this study integrates these approaches into a unified evolutionary perspective.
Design: This study employs a conceptual research design based on an integrative literature review. Relevant studies on software development paradigms, software reuse, low-code/no-code platforms, generative AI, vibe coding, and agentic software engineering were systematically analyzed and synthesized to develop a conceptual framework describing the evolution of application development. The proposed framework was subsequently analyzed through three analytical dimensions: development abstraction, developer role transformation, and degree of automation.
Findings: The study identifies five dominant application development paradigms: Manual Coding, Framework-based Development, Low-Code Development, AI-assisted Development, and Agentic Development. The findings indicate that application development has progressively evolved toward higher levels of development abstraction, increasing software engineering automation, and continuous transformation of developer roles from software implementers to AI collaborators and orchestrators. The study further argues that these paradigms should be viewed as complementary and evolutionary rather than mutually exclusive, as contemporary software development frequently combines multiple paradigms within a single project.
Originality/Value: This study contributes a unified conceptual framework that integrates previously fragmented software development paradigms into a coherent evolutionary model. The framework provides a theoretical foundation for future research in software engineering and offers practical implications for redesigning application development curricula and preparing developers for increasingly AI-driven and agentic software engineering environments.