AI-Driven E-Commerce Automation Using Blazor Web Framework

Abstract

This thesis presents the development of the Tanui Watch Store, a full-stack luxury e-commerce platform built using the Microsoft Blazor WebAssembly framework. The research addresses critical challenges in digital retail, specifically focusing on reducing response latency for complex technical queries and overcoming regional payment friction in emerging markets like Kenya. By implementing a Retrieval-Augmented Generation (RAG) pipeline with Qdrant and OpenAI, the system provides factually grounded, human-like AI responses based on trusted company documentation. Additionally, the platform integrates the Safaricom Daraja API to support localized M-PESA STK Push payments, facilitating a seamless one-click transaction process for mobile money users. The project utilizes an N-tier layered architecture with a .NET Core backend, leveraging SignalR for real-time status updates and Azure for secure cloud deployment.