Identification of the correct replacement parts was forming a bottleneck in the process and often resulted in misidentified items. The existing process was a very time consuming and manual one involving a discussion with the customer, browsing through thick spare part catalogs, and always carrying the risk of ordering a wrong component. We automated this entire process with machine vision technology and integrated the identification process directly into a new e-commerce platform.
Through a WhatsApp integrated service, you can quickly and easily take a photo of the broken component and send it directly to Green Master. Once in our system, our machine vision application uses identifiable data points from the submitted imagery to match it to an item in the Green Master catalog of parts with a far higher degree of confidence than the manual process.
Once an item is identified the system will automatically reply with a link to the identified part within the Green Master e-commerce platform, where the user can directly purchase the replacement or new component in confidence that their order will match their need.
The efficiencies created by the system impacted not just Green Master as a company, but also their customers. The reduction of manual consultation and resolving order issues meant vast time savings. The reduction in misidentified parts meant more satisfaction for their customers whose confidence in the product, and Green Master, greatly improved as a result.
• Product Strategy
• User Research & Testing
• Service Design
• Technical Architecture Planning
• Machine Vision (BigTransfer)
• AI Engineering & Machine Learning
• Cloud Services Technologies (AWS)
• API Development
Data Scientist, Programmer at Finlabs
We developed this with some of the newest technical solutions available, so this was a very exciting project for us. I think this demonstrates how quickly we can adapt to new technologies and create the best value possible for the customer.