About

• Python-based app that creates realistic, customizable mock inventory datasets for testing, planning, and analysis.

• Unlike traditional mock-data tools that rely on fixed categories & rigid bin structures, this app leverages AI to dynamically generate categories, subcategories, & bin-location conventions based on a user’s requirements.

Methodology

• Uses LLMs via OpenAI to interpret short business descriptions & return structured JSON schemas for categories & bin rules.

• Synthetic data generation powered by Faker & NumPy, ensuring realistic item names, stock levels, reorder points, unit costs & timestamps.

• Bin location codes are generated flexibly, supporting aisles, zones, shelves & custom formats.

• Data is output in CSV format for compatibility with Excel, Power BI, Python & ERP systems.

• GUI built with Tkinter for accessibility & user-friendly interaction.

Technologies Used

• Python

• Pandas & NumPy (data handling)

• Faker (synthetic but realistic text & date generation)

• Tkinter (desktop GUI)

• LangChain + OpenAI API (AI-powered schema inference)

Current State Gap Identification

Traditional inventory mock-data tools and scripts typically:

• Rely on hardcoded categories & bin conventions, making them inflexible.

• Require significant manual setup to simulate realistic stock datasets.

This project fills the gap by providing domain-adaptive, AI-driven dataset generation that adjusts to user context.

Specific Objectives

1. Provide an easy-to-use tool for generating large-scale mock inventory datasets.

2. Enable AI-driven schema inference for categories, subcategories & bin locations.

3. Improve efficiency for testing, training, and validating inventory management systems, dashboards, and ERP workflows.

4. Ensure outputs are scalable (100–100,000+ rows) and interoperable with industry tools.

5. Offer a lightweight solution, accessible to technical and non-technical users.

Significance & Implications of Work

• For developers: Accelerates ERP & dashboard testing by providing realistic data without needing access to sensitive company datasets.

• For analysts: Supports stress-testing of forecasting, stock optimization, and procurement models with flexible input data.

• For organizations: Reduces reliance on live production data, improving data privacy & compliance during development.

Clear Prerequisites & Requirements

• Python 3.9+
• OpenAI API key (for AI-driven schema inference; fallback available if AI is not accessible)
• Packages: pandas, numpy, faker, tkinter, langchain-openai