Imagine what you could do with meldCX's Machine Learning on Microsoft’s Azure Cloud
meldCX, together with its affiliate company AOPEN, will highlight the latest in machine learning (ML) and edge processing at Computex Taipei 2018. This game-changing Azure solution provides a real glimpse of the future of artificial intelligence (AI) and cloud computing for commercial applications.
Developed for Prolife Foods, experts in the marketing and managing of self selection foods within Australia and NZ’s leading supermarket chains, meldCX’s product recognition technology, m-intelligence, is a leading AI Initiative that empowers a company to cultivate and teach their own training models.
M-intelligence adopts an edge processing methodology which allows transactional speeds offline while being updated and educated online. Images are captured by a human trainer, uploaded, analysed and the reference data is stored in Azure Cloud Storage. Training of the m-intelligence models are done using the GPU enabled N-series virtual machines on Azure.
The use of Azure N-series services has significantly advanced the project: making it possible to train a single model in a few hours verses 3 days.
Ian Jackson, General Manager of Prolife Foods explains the capabilities of meldCX’s ML technology to simplify a complicated supply chain environment: “By applying advanced computer vision algorithms to a camera attached to a scale, the system can see and understand what is going on at the point of transaction and track every item.
“This project was very complex but working with meldCX and AOPEN we have eliminated development blockers, which will have a very meaningful impact on our bottom line.
“We can apply this machine learning model across our business and integrate with our supply chain to optimize how we utilize end-to-end information.”
Stephen Borg, Co-founder, and CEO of meldCX, says that meldCX has made it more accessible for companies wanting to develop a solution on AI and most importantly, seamlessly integrate this technology into multiple applications.
“We have created more than a stand-alone, isolated solution. We see m-intelligence being applied in many ways including Point-of-Sale (POS), behavioural learning, self check-out, self service kiosk and assisted check out. Our ML solution is also a broad set of tools that not only identify but interact with other devices, systems and commercial peripherals to ensure a comprehensive end to end solution.
“In this case, it allows Prolife to train their own models without writing a single line of code. When coupled with AOPEN’s 22” commercial touch device, meldCX is able to provide a robust solution that meets Prolife’s business requirements and can be deployed and managed at scale from a central console”.
MeldCX collaborated with Microsoft to develop and host the solution in Azure, and lock-down devices for additional security and stability during software OS upgrades. Endorsed technically by the Azure team, the meldCX solution is one of the first Azure based sophisticated machine learning solutions that is ready for commercialisation.
“We are pleased that meldCX is utilizing the powerful graphics processing capabilities of Azure N-series GPU enabled virtual machines as the foundation for M-Intelligence,” said Shirley Strachan, General Manager, IoT, Microsoft. “The meldCX solution will help businesses build and use innovative AI solutions to deliver entirely new customer experiences in physical environments.”