The new Google headquarters in the U.K. (Photographer: Chris Ratcliffe/Bloomberg)

Google Pitches AI Image-Detection Tool to Lure Cloud Clients

(Bloomberg) -- Alphabet Inc.’s Google is introducing a new product that lets data-center customers tap the search engine’s artificial-intelligence arsenal, part of the company’s effort to stand out in the competitive cloud-computing market.

The product, called Cloud AutoML, is aimed at companies with limited experience in AI. It uses a business’s data to automatically generate machine-learning models -- computing systems that can crunch and parse huge swaths of information. Google is starting with vision, allowing companies to classify images more effectively, but plans to release the tool for analyzing text, speech and video.

"It’s dead simple," Rajen Sheth, a senior director for Google Cloud, said on a phone call with reporters. "Somebody with very little expertise can go and create something that solves their very specific business problem." Urban Outfitters Inc. and Walt Disney Co. both said they’re using the Google product to spot trends and structure their e-commerce sites. Google declined to share pricing details.

Access to machine learning is a major selling point for Google’s cloud business, which is third in the market behind Amazon.com Inc. and Microsoft Corp. Google first introduced its AutoML feature last May, as an internal research project, touting it as a novel machine-learning system that could, in effect, create other AI systems on its own. It’s designed to perform complex operations with little to no advanced software programming. Google hopes to sign up more cloud customers but it’s opening the product to any company interested, a spokesman said.

Research firm KeyBanc wrote last week that Google’s share in the cloud grew to 12 percent in the fourth quarter from 10 percent a year earlier, estimating Google’s sales in the division at $3.9 billion on an annualized basis. KeyBanc pegged Amazon’s cloud revenue annual run-rate at $19.8 billion and Microsoft at $6.4 billion.

©2018 Bloomberg L.P.