Code - Architect AI: Enterprise Documentation for Repository Analysis and AI-Driven Insights

Abstract Imagine grappling with a sprawling codebase where hidden complexities threaten to derail your next release—that's the reality for many developers. This paper introduces "AI Architecture Pro," an intuitive tool designed to transform code analysis from a tedious chore into a proactive ally. By harnessing static analysis, graph theory, machine learning (ML), and large language models (LLMs), it effortlessly dissects polyglot codebases in Python, JavaScript, and Java, unveiling architectural insights that traditional tools overlook. At its core, the tool—built as a user-friendly Streamlit web app—employs libraries like NetworkX for dependency graphs, Radon for precise metrics, and Groq for intelligent AI chat. Developers can explore interactive 3D visualizations of "code cities," where skyscrapers represent complexity, and receive ML-driven predictions on refactoring risks. It also generates clustered architecture maps and radar charts for technical debt assessment, all while supporting GitHub cloning, local paths, or pasted code for flexibility. Evaluations on real-world repositories show it pinpointing high-risk functions with 85% accuracy and boosting maintainability scores by up to 20%. Ultimately, "Code Architect AI" bridges the divide between raw metrics and actionable wisdom, empowering teams to refactor smarter and build more resilient software. Keywords:- code analysis, Software Architecture,Machine Learning, Static Analysis, Visualizations, AI -Assisted Development, Dependency Graph, Predictive Maintenance.