Design and Implementation of AI-Driven Microservices Architecture for Scalable Machine Learning Workloads

Abstract

This paper explores the design and implementation of a microservices-based architecture tailored for scalable machine learning (ML) workloads enhanced by artificial intelligence (AI). It discusses how microservices facilitate flexible deployment, scalability, and real-time model updates while allowing AI components to dynamically manage workload distribution. Two diagrams are included to illustrate the architecture and AI-driven orchestration, alongside two tables that compare legacy vs. microservices-based ML systems and performance metrics. This architecture is especially suited for cloud-native, continuous-integration environments, where AI drives decisions about resource scaling and service orchestration.