The NGS Inventory Management System is a full stack project I developed using React, Express.js, and libraries like AG Grid, AG Charts, and Shadcn.

It was created for Ng Guan Seng Woodworking Industrial (NGS) to meet their specific business needs. One of the standout features was an email scanner that automatically updated inventory data from supplier emails.

That feature wasn’t common in existing systems, so I had to design both the UI and backend logic from scratch. Parsing both structured and unstructured emails required a mix of rule based and pattern matching approaches.

The project came with its fair share of challenges. Managing complex datasets and maintaining performance under heavy load required balancing clean code with real world efficiency.

Besides tracking inventory, the system also handled purchase orders, low stock alerts, and analytics dashboards to help NGS monitor operations. The backend was also structured to allow future integrations with supplier APIs.

This was also where I experimented with an agent style AI chatbot that could perform data analysis on command. Users could ask it to generate insights, and it would automatically query, process, and visualize the data in customizable dashboards.

Working on this system showed me how crucial it is to understand a client’s workflow deeply and iterate closely with them. It strengthened my ability to build enterprise grade software that’s robust and practical.

In the end, the system significantly reduced manual work and improved data accuracy for NGS. Even though it wasn’t deployed commercially, the project gave me valuable experience in handling real world business software challenges.