AI-Driven Decision Support Systems for Software Architecture: A Framework for Intelligent Design Decision-Making (2025)

Abstract

Software architecture decision-making is a critical phase in the software development lifecycle, often constrained by time, complexity, and uncertainty. As software systems grow in scale and dynamism, architects require intelligent tools that can assist in evaluating architectural alternatives, predicting quality trade-offs, and automating design suggestions. This paper explores the integration of Artificial Intelligence (AI) into the software architecture decision-making process. We review existing AI-driven architectural tools, classify relevant AI techniques (including expert systems, machine learning, and large language models), and propose a conceptual framework for an AI-based Architecture Decision Support System (AIDSS). The proposed system aims to enhance decision quality by learning from historical design data, recommending optimal patterns, and ensuring traceability. We evaluate the framework against current challenges such as explainability, context-awareness, and evolving system requirements. A visual model of the system architecture is presented, along with example use cases. This paper aims to provide both theoretical insights and practical directions for implementing AI-assisted decision-making in real-world architectural practices.