Caterlynx

Client: Caterlynx Glass

Contracted by: Caterlynx Glass
Project: Intelligent Online Store
Overview: An online cater ware store
Project link: www.caterlynxglass.co.za

Tags: Deep Learning, Machine Vision, C++, Python, Robotic Process Automation

Caterlynx Glass is a large supplier of glassware and catering accessories to the Restaurant industry. They needed an Intelligent online store that used Robotic Process Automation to automate some of their inventory assignment and tracking. They also needed mobile applications for their customers to easily place custom orders and manage their accounts

We built the store with a customized dashboard for our client. Inventory loading is automated by loading images and descriptions from supplier catalogues. We use machine vision and machine learning techniques to detect what type of image is being viewed to place it in the correct category. Inventory is also loaded from spreadsheets and thousands of downloaded images. This process used to take weeks for our client, but now it takes a few minutes

Mobile applications were created to enable customers to easily access the database of products as well as mange their accounts.

Technologies Used

Sales site
The sales site is created in PHP with a modified popular shopping cart framework. The API used by the mobile applications is also written purely in PHP

Machine Vision and Machine Learning
We use OpenCV and an application written in C++ for image preprocessing and feature detection. Google’s Tensor Flow is used in a Python Application for image classification with a Deep Learning Neural Network