Application of AI in E-Commerce for Startups with Limited Data: A Lean Startup Approach to Demand Forecasting and Inventory Optimization

Elizondo De La Garza, Ana Paula (2025) Application of AI in E-Commerce for Startups with Limited Data: A Lean Startup Approach to Demand Forecasting and Inventory Optimization. Masters thesis, Dublin, National College of Ireland.

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

Abstract

Startups in the e-commerce sector face critical operational challenges due to data-scarcity, inventory volatility and limited resources, which significantly affects demand forecasting accuracy and efficient inventory planning. This research explores how AI-based solutions can benefit startups in these specific contexts by designing and evaluating a lightweight, startup-ready forecasting system. This study is tested with real transactional data from a Mexican perfumes e-commerce startup, LeBorêt, and compared four predictive models: SARIMAX with bootstrapping, baseline LSTM, enhanced LSTM with data augmentation and transfer learning, and XGBoost with fine-tunned features. Evaluation results shows that while LSTM-models offer strong adaptability, SARIMAX combined with bootstrapping provided the most operationally significant balance between accuracy levels and flexibility, achieving 93.3% coverage in a predictive interval. The selected forecasting method was integrated into a minimum viable product (MVP), “COSMOS”, following the Lean Startup methodology focusing on validated learning by co-designing this interface with the startup’s founder. The system offers dynamic forecasts, smart inventory alerts and various planning tools. This project contributes to AI democratization, and offers a validated, replicable methodology for AI adoption in early-stage startups with resource-constrained situations.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Del Rosal, Victor 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 > HF Commerce > Electronic Commerce H Social Sciences > HD Industries. Land use. Labor > New Business Enterprises
Divisions:	School of Computing > Master of Science in Artificial Intelligence for Business
Depositing User:	Ciara O'Brien
Date Deposited:	24 Jun 2026 11:23
Last Modified:	24 Jun 2026 11:23
URI:	https://norma.ncirl.ie/id/eprint/9400

Actions (login required)

View Item