Smart Retail System : AI Analyzer Module

Tan, Jia Seng (2025) Smart Retail System : AI Analyzer Module. Final Year Project (Bachelor), Tunku Abdul Rahman of Management and Technology.

Text RSW_Tan Jia Seng_Fulltext.pdf Restricted to Registered users only Download (2MB)

Abstract

The generation of reports has evolved significantly over time, particularly with advancements in technology. Tasks that were once performed manually have gradually shifted to automated processes, offering greater convenience and efficiency. Traditional methods of report generation required humans to manually arrange and input data into systems, which increased the risk of human error such as entering incorrect data or omitting important information. These inefficiencies can lead to reduced accuracy in reports, resulting in ambiguity in decision-making and potential financial losses. With the implementation of AI technology, where data is collected and processed automatically for report generation, systems are expected to become more accurate and efficient in managing reports. In this part of the project, user identification and AI Analyzer module is mainly covered. The covered modules include User Management Module, Sales Trend Prediction Report Modules, Customer’s Satisfaction Report Module and Periodic Report Module and Report Database Management Module which supports the three modules stated earlier. The system will be developed using HTML, CSS, JavaScript along with chart.js and plotly.js for the front-end development. On the other hand, Flask and Python are used to develop the back-end of the system which server as the system infrastructure. The Agile Software Development Model is used due to the limited time and project size and to determine the functional and non-functional requirements of the system. Testing was carried out following the flow of unit testing, integration testing, system testing and lastly acceptance testing. Ultimately, the system proved to increases the efficiency, effectiveness and easier generation for reports. Although so, limitations such as the accuracy of the prediction reports could be further improved as if an actual retail database is provided.