×

Amazon has a tech team focused on helping customers find clothes that fit perfectly

  • An engineer at Amazon posted a LinkedIn description that suggests Amazon is researching ways to help customers buy clothes that fit.
  • Some of the challenges described in the profile include hardware and robotics for measuring items, and research to build 3D models and predict how items will fit.
Amazon CEO Jeff Bezos discusses the Fire smartphone at a news conference in Seattle, June 18, 2014.
Jason Redmond | Reuters
Amazon CEO Jeff Bezos discusses the Fire smartphone at a news conference in Seattle, June 18, 2014.

Amazon has been quietly assembling a team of people to build technology that could one day help you find shoes and clothes that fit just right.

Should the technology become available, it would be another way Amazon is reinventing retail with machine learning.

Sam Dwarakanath, a senior manager of software engineering who previously worked on the Amazon Go smart concept convenience store, described the effort at a high level a few months ago on his LinkedIn profile.

"As of September 2016 I am building and leading a technical team that is focused on making it easy for customers to discover apparel and shoes that appeal to them, including finding the perfect fit," Dwarakanath wrote. Some of the challenges described in the profile include hardware and robotics for measuring items, and research to build 3D models and predict how items will fit.

Given that description, it sounds like Amazon wants to cut down on its customers returning ill-fitting products.

The final technology could involve taking a picture of a product with your phone and then showing how different sizes of the product would fit on you.

Amazon already lets people order products just by speaking through its Alexa virtual assistant, and with Amazon's main mobile app, people can identify and then order products by scanning objects with image recognition. Similar visual search technology is available from Pinterest and Alphabet's Google.

Amazon did not immediately respond to a request for comment.