A Resilient Microservices Architecture for AI-Powered Academic Assistant: Integrating Spring AI, RAG, and MCP Patterns

This record provides the preprint and associated resources for the study entitled: "A Resilient Microservices Architecture for AI-Powered Academic Assistant: Integrating Spring AI, RAG, and MCP Patterns" This work presents the design, implementation, and evaluation of a distributed system for an AI-powered academic assistant, developed using a microservices architecture based on Spring Cloud. The system integrates Retrieval-Augmented Generation (RAG) with Spring AI to enable domain-specific question answering grounded in structured data about academic faculty members. The architecture incorporates multiple modern software engineering and AI integration patterns, including:- Retrieval-Augmented Generation (RAG) for contextualized responses- Model Context Protocol (MCP) for structured interaction with internal and external data sources- Serverless processing using Spring Cloud Function- Resilience patterns (Circuit Breaker, Retry, Rate Limiter, Time Limiter) implemented via Resilience4j- Full observability using Prometheus, Grafana, and Zipkin The system was evaluated through load testing using Apache JMeter, demonstrating robustness under failure scenarios and the effectiveness of resilience mechanisms in preventing cascading failures. Source code:https://github.com/DougFelipe/rag-academic-assistant Licensing:- This manuscript is licensed under Creative Commons Attribution 4.0 International (CC BY 4.0)- The associated software is distributed under the Apache License 2.0