Shopmicro Ai microservices ecommerce platform

Abstract

The rapid expansion of online commerce requires platforms that remain responsive, secure, maintainable, and personalized as user traffic and business functions grow. This paper presents ShopMicro, an AI-enabled e-commerce platform developed with a microservices architecture. A React and TypeScript frontend communicates through a centralized API Gateway with independently deployable Authentication, Product, Order, Payment, and Inventory services implemented using Node.js and Express.js. Each service maintains an isolated PostgreSQL database, reducing coupling and improving fault isolation. The platform integrates the Groq large-language-model API as an intelligent shopping assistant and Pinecone vector search for semantic product discovery beyond exact keyword matching. JWT-based authentication, bcrypt password hashing, protected routes, and role-based access control secure customer and administrator workflows. Experimental observations reported by the project show API Gateway response times of approximately 80–150 ms, product-service responses of 100–180 ms, order processing of 350–600 ms, payment processing of 300–500 ms, inventory updates below 100 ms, and service availability of 99.8%. The results indicate that the proposed architecture provides a practical foundation for scalable, cloud-ready, and intelligent ecommerce applications while preserving modularity and independent service evolution.