This research introduces MicroPAD, a prototype tool that automatically detects microservice pattern instances by analyzing Infrastructure-as-Code artifacts with large language models.

• Addresses the critical challenge of knowledge vaporization in software architecture
• Shifts analysis from traditional source code to Infrastructure-as-Code artifacts
• Leverages LLMs to identify and document architectural patterns automatically
• Helps preserve valuable architectural knowledge with minimal manual effort

For engineering teams, this approach offers a practical solution to maintain accurate architecture documentation while reducing the documentation burden on developers.

Automated Microservice Pattern Instance Detection Using Infrastructure-as-Code Artifacts and Large Language Models