Reproduce “Bridging Requirements and Architecture: Multi-Agent Orchestration with External Knowledge and Hierarchical Memory” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Ruiyin Li, Yiran Zhang, Xiyu Zhou, Yangxiao Cai, Peng Liang, Weisong Sun, Jifeng Xuan, Zhi Jin and 1 more
MAAD, a multi-agent framework, automates software architecture design by transforming requirements into comprehensive, multi-view architectural blueprints, leveraging external knowledge and hierarchical memory. The method generates structured and consistent artifacts, demonstrating improved modularity and cohesion compared to baseline systems, and adapts design rigor based on the underlying large language model's capabilities.
Key insights
- Orchestrated specialized agents, supported by external knowledge and hierarchical memory, effectively bridge the gap between high-level requirements and detailed architectural designs.
- The infusion of external knowledge via RAG promotes standard-compliant architectural structures with clearer abstraction layers, enhancing traceability and role definition.
- The quality and detail of the generated architecture artifacts are highly dependent on the reasoning capacity of the foundational LLM used within the framework.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.