The Smart Inventory and Demand Forecasting System is a web-based application engineered to address the operational inefficiencies that pervade manual inventory management in small and medium-sized enterprises (SMEs). Conventional approaches based on spreadsheets, paper ledgers and disconnected accounting tools commonly produce data inconsistencies, delayed decision-making, stock shortages and surplus accumulation, which translate into measurable financial losses and customer dissatisfaction. The proposed platform introduces a unified digital workspace in which administrators and inventory managers can upload product data from heterogeneous file formatsincluding Microsoft Excel spreadsheets, Portable Document Format files and Microsoft Word documentsand have the system automatically extract, normalize, categorize and persist the data to a relational database. Inventory movements are tracked across their full life cycle, distinguishing incoming stock transactions (goods received from distributors) from outgoing stock transactions (goods sold to customers). The system is implemented in Python using the Django web framework, with pandas for tabular manipulation, pdfplumber for PDF table extraction and python-docx for Word document parsing; SQLite serves as the default backend, and the application follows Django's Model-View-Template (MVT) architectural pattern. A dynamic dashboard reports total product counts, current stock levels, weekly and monthly sales, low-stock alerts, highdemand items and category-wise breakdowns, while a report-generation module produces professional PDF and Word documents that support audit, procurement and customer-facing communication. Together these capabilities reduce manual effort, improve data accuracy and provide data-driven insight into demand patterns. The resulting system constitutes a practical and deployable solution that allows SMEs to modernize their inventory operations without incurring the licensing or implementation cost of commercial enterprise resource planning suites.
Chandana et al. (Thu,) studied this question.