Google Cloud AutoML lets business play with AI without coding

Artificial intelligence and machine learning are the buzzwords of today and everyone and anyone, especially businesses, want in. But while the results of using AI and ML seem magical, getting there is anything but. At the very least, you need to train a computer and build a model for a specific business case, which often requires not just engineers but also time to build up that model. Enter Google's latest machine learning enterprise, Cloud AutoML, designed to make building such models as easy as dragging and dropping files.

AI and machine learning don't happen by magic and are, in fact, resource intensive on all fronts. You need programmers and engineers familiar with the science to develop the software for it. You need datasets to feed into the machine. And you need time to build up that machine learning model. It's a problem for companies both big and small alike.

Google's new Cloud AutoML takes out 2 out of 3 parts of that equation. You no longer need access to human engineers nor do you need to spend so much time on the process. It uses learning methods to simplify the process for businesses. All they need to do is send files, wait, and then test the predictions.

The first product to come out of this new tool is Cloud AutoML Vision. As the name suggests, it is designed to build image recognition models. Users simply have to drag and drop images to build a model specific to a use case, like identifying clouds types.

Disney, for example, uses it to annotate characters and products to help make searching easier for customers. The Zoological Society of London uses it to save time, money, and human resources from having to manually analyze and annotate the millions of pictures captured by their camera traps in the wild. AutoML Vision isn't yet available to all businesses but you can try requesting access if you're interested.

SOURCE: Google