AI-Powered Predictive Dashboard for Retail Decision-Making

Pinto Elchiver, Alfonso Armando (2025) AI-Powered Predictive Dashboard for Retail Decision-Making. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (1MB) | Preview
Preview	PDF (Configuration Manual) Download (2MB) | Preview

Abstract

The growing volume of sales, inventory, and supplier data presents an opportunity for artificial intelligence (AI) to transform retail decision-making. However, many small and medium-sized enterprises lack the tools and expertise to apply predictive analytics effectively. This research addresses that challenge by designing and evaluating an AI-powered dashboard that merges advanced forecasting models with a simple, role-adapted interface. Using the publicly available “Warehouse and Retail Sales” dataset (307,647 records), three models — Prophet, XGBoost, and Long Short-Term Memory (LSTM) — were implemented to forecast demand, highlight inventory risks, and suggest optimal promotion timing. An adapted CRISP-DM process, extended with stakeholder interviews and usability testing, ensured alignment between technical capability and business needs.

Findings show that the dashboard delivered accurate forecasts, with LSTM achieving the highest overall accuracy, XGBoost performing consistently well across diverse product categories, and Prophet offering the clearest interpretability for business users. Equally important, these forecasts were presented in a way that managers could easily understand and act upon. Stakeholders described the dashboard as intuitive, relevant to their daily tasks, and capable of supporting timely, informed decisions. By combining technical robustness with practical usability, the prototype bridges the gap between AI potential and everyday retail practice, with future work aimed at live deployment and the inclusion of broader business variables.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Haque, Rejwanul UNSPECIFIED
Subjects:	Q Science > QH Natural history > QH301 Biology > Methods of research. Technique. Experimental biology > Data processing. Bioinformatics > Artificial intelligence Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Artificial intelligence H Social Sciences > HD Industries. Land use. Labor > Specific Industries > Retail Industry
Divisions:	School of Computing > Master of Science in Artificial Intelligence for Business
Depositing User:	Ciara O'Brien
Date Deposited:	24 Aug 2026 11:02
Last Modified:	24 Aug 2026 11:02
URI:	https://norma.ncirl.ie/id/eprint/9593

Actions (login required)

View Item