An LLM-assisted approach to designing software architectures using ADD

AI-generated keywords: Software Architecture Design LLMs ADD Approach Collaboration Human Oversight

AI-generated Key Points

The license of the paper does not allow us to build upon its content and the key points are generated using the paper metadata rather than the full article.

  • Software architecture design is complex and iterative, relying on expert judgment.
  • The authors propose a method that uses Large Language Models (LLMs) to assist in software architecture design through the Attribute-Driven Design (ADD) approach.
  • The method involves equipping an LLM with a detailed description of ADD, an architect persona, and a structured iteration plan for collaborative generation of architecture artifacts between the LLM and a human architect.
  • Rigorous case studies validate the effectiveness of the LLM-assisted ADD process by comparing generated designs against established solutions and evaluating them with professional architects.
  • Architectures produced through this collaborative approach closely align with proven solutions while partially meeting architectural drivers, highlighting both potential and current limitations of employing LLMs in architecture design.
  • Human oversight and iterative refinement are crucial when integrating LLMs into software architecture design to achieve optimal outcomes.
  • The research emphasizes how advanced technologies like LLMs can enhance traditional software architecture design processes when combined with established methodologies like ADD.
Also access our AI generated: Comprehensive summary, Lay summary, Blog-like article; or ask questions about this paper to our AI assistant.

Authors: Humberto Cervantes, Rick Kazman, Yuanfang Cai

30 pages, 12 figures, 7 tables

Abstract: Designing effective software architectures is a complex, iterative process that traditionally relies on expert judgment. This paper proposes an approach for Large Language Model (LLM)-assisted software architecture design using the Attribute-Driven Design (ADD) method. By providing an LLM with an explicit description of ADD, an architect persona, and a structured iteration plan, our method guides the LLM to collaboratively produce architecture artifacts with a human architect. We validate the approach through case studies, comparing generated designs against proven solutions and evaluating them with professional architects. Results show that our LLM-assisted ADD process can generate architectures closely aligned with established solutions and partially satisfying architectural drivers, highlighting both the promise and current limitations of using LLMs in architecture design. Our findings emphasize the importance of human oversight and iterative refinement when leveraging LLMs in this domain.

Submitted to arXiv on 27 Jun. 2025

Ask questions about this paper to our AI assistant

You can also chat with multiple papers at once here.

The license of the paper does not allow us to build upon its content and the AI assistant only knows about the paper metadata rather than the full article.

AI assistant instructions?

Results of the summarizing process for the arXiv paper: 2506.22688v1

This paper's license doesn't allow us to build upon its content and the summarizing process is here made with the paper's metadata rather than the article.

In their paper titled "An LLM-assisted approach to designing software architectures using ADD," authors Humberto Cervantes, Rick Kazman, and Yuanfang Cai delve into the intricacies of software architecture design. They highlight the complexity and iterative nature of this process, typically reliant on expert judgment. The authors propose a novel method that leverages Large Language Models (LLMs) to assist in software architecture design through the Attribute-Driven Design (ADD) approach. By equipping an LLM with a detailed description of ADD, an architect persona, and a structured iteration plan, the authors' method facilitates collaborative generation of architecture artifacts between the LLM and a human architect. Through rigorous case studies, where generated designs are compared against established solutions and evaluated by professional architects, the effectiveness of the LLM-assisted ADD process is validated. The results demonstrate that architectures produced through this collaborative approach closely align with proven solutions while partially meeting architectural drivers. This underscores both the potential and current limitations of employing LLMs in architecture design. The findings underscore the critical role of human oversight and iterative refinement when integrating LLMs into this domain. Overall, Cervantes, Kazman, and Cai's research sheds light on how advanced technologies like LLMs can enhance traditional software architecture design processes when used in conjunction with established methodologies like ADD. Their work emphasizes the importance of maintaining a balance between automated assistance and human expertise for optimal outcomes in software architecture development.
Created on 22 Apr. 2026

Assess the quality of the AI-generated content by voting

Score: 0

Why do we need votes?

Votes are used to determine whether we need to re-run our summarizing tools. If the count reaches -10, our tools can be restarted.

Similar papers summarized with our AI tools

Navigate through even more similar papers through a

tree representation

Look for similar papers (in beta version)

By clicking on the button above, our algorithm will scan all papers in our database to find the closest based on the contents of the full papers and not just on metadata. Please note that it only works for papers that we have generated summaries for and you can rerun it from time to time to get a more accurate result while our database grows.

Disclaimer: The AI-based summarization tool and virtual assistant provided on this website may not always provide accurate and complete summaries or responses. We encourage you to carefully review and evaluate the generated content to ensure its quality and relevance to your needs.