Client Name

Mr. Shehryar Teli

Faculty Advisor

Dr. Mohammad Kamran Mumtaz, Assistant Professor, Department of Management, Institute of Business Administration

SBS Thought Leadership Areas

Entrepreneurship and Innovation

SBS Thought Leadership Area Justification

  • The project introduced custom-built forecasting models and a Python-based dashboard tailored for an SME (AZ Textiles), showcasing technological innovation in a traditional industry.

  • It empowered a mid-sized business to move toward data-driven decision-making, a hallmark of entrepreneurial transformation.

Aligned SDGs

GOAL 9: Industry, Innovation and Infrastructure

Aligned SDGs Justification

This Experiential Learning Project directly supports SDG 9 by:

  • Digitizing planning systems in a mid-sized manufacturing SME.

  • Replacing informal, intuition-based forecasting with data-driven forecasting and inventory optimization.

  • Promoting technological innovation through a Python-based dashboard built on Flask and Dash.

  • Strengthening operational infrastructure via automation, cost reduction, and improved stock planning.

  • Enabling scalable, industry-relevant innovation that can be replicated in similar SMEs across Pakistan and other developing countries.

The project helps transition traditional manufacturers toward smart, resilient, and innovative supply chains, making it an ideal match for SDG 9.

NDA

No

Abstract

This Experiential Learning Project (ELP) was undertaken to address the recurring issue of stockouts at AZ Textiles, a Pakistan-based exporter of textile-based promotional products. These stock outs were leading to lost orders and declining customer satisfaction - particularly in the U.S. wholesale market. The project’s primary objective was to analyze historical sales and inventory data, develop a reliable demand forecasting model, and design an automated inventory and manufacturing planning system. The study focused specifically on QTB bags, comprising 26 SKUs, to create a targeted and effective solution.

The project employed a structured approach combining data analytics, forecasting techniques, and interactive dashboard development. Sales and inventory data were cleaned and processed, followed by an exploratory analysis to identify seasonal trends, shifts in SKU demand, and year-over-year patterns. An interactive dashboard was developed using Python (Flask and Dash) to visualize these insights and support decision-making at the management level.

For forecasting, multiple models were tested. Ultimately, a log-linear regression model was selected for its interpretability and robustness, achieving a Mean Absolute Percentage Error (MAPE) of approximately 11% when back tested on yearly data. Forecasts were conducted at the category level, and subsequently disaggregated to the SKU and month level using product and monthly weights derived from historical contributions. This ratio-based methodology offered a practical solution to data limitations and allowed for seasonally adjusted, SKU-specific forecasts.

In parallel, the team implemented the Clark and Scarf Inventory Optimization Model to convert demand forecasts into operational inventory plans. This model calculated optimal order quantities, reorder points, and safety stock for each SKU across a five-year horizon (2025–2030), accounting for lead time and cost constraints.

The key deliverables included a validated forecasting model, a dynamic dashboard for operational visibility, and a set of actionable recommendations for improving AZ Textiles’ inventory management. The project aligns with Sustainable Development Goal (SDG) 9: Industry, Innovation, and Infrastructure, by encouraging the digital transformation of manufacturing SMEs through the use of data-driven planning systems.

The outcomes contribute to enhanced operational resilience, reduction in stockouts, and lay the groundwork for scalable, technology-enabled supply chain optimization at AZ Textiles.

Document Type

Restricted Access

Document Name for Citation

Experiential Learning Project

Notes

This project was developed in collaboration with AZ Textiles and supervised by Dr. Mohammad Kamran Mumtaz as part of the BBA Experiential Learning Program at IBA Karachi. The forecasting methodology was validated through client meetings, and the dashboard prototype was reviewed and appreciated by the company’s COO.

Download

Available for download on Monday, December 15, 2025

Share

COinS