Watch all of the Transform 2020 periods on-demand proper right here.
While eBay may have used any variety of current AI platforms to reinforce its varied merchandise, the corporate as an alternative elected to construct its personal AI system — dubbed Krylov — in-house and make it open supply for anybody to make use of. That determination seems to be paying off.
The San Jose-based firm has made no secret of its AI ambitions over the previous 4 years, hoovering up technical expertise by way of acquisitions and launching myriad automated instruments to identify bank card fraud, enhance product listings, and produce different nifty options to patrons and sellers. Late final 12 months, eBay additionally relaunched a standalone vehicle-focused Motors app, with AI and automation at its core.
Such is the pervasiveness of AI throughout eBay in 2020 that the corporate’s chief expertise officer, Mazen Rawashdeh, says AI is now an “ecosystem” on the firm.
“I don’t call AI a product — I call it an ecosystem,” Rawashdeh stated throughout a chat at Transform 2020 in the present day. “Because it’s moved into almost everything we serve for our customers, for everything we do internally to manage our platform. We use it for many areas from image search to personalization, recommendations, price guidance, fraud detection, and shipping estimates. So it is absolutely an ecosystem for us.”
These varied AI use circumstances assist for example why eBay bothered to construct Krylov within the first place: It all comes right down to scalability and enabling a “shared model” between groups. According to Rawashdeh, Krylov has helped scale AI throughout the corporate and provides all of the enterprise items easy accessibility to pure language processing (NLP), pc imaginative and prescient, and machine studying smarts, whereas additionally bypassing the necessity to sew collectively elements from varied different AI platforms.
“So eBay had some form of AI capabilities three or four years ago, but it was disjointed, and it was a siloed model,” he defined. “It was not a capability that — end-to-end — takes the process of data, of compute, of shared model, and allows our researchers and data scientists to go to one place and run their model at scale.” Thus, Krylov successfully permits eBay to centralize its AI, saving particular person researcher and improvement groups from working resource-intensive fashions individually.
The COVID-19 disaster has additionally shone a light-weight on a number of the advantages of working a centralized end-to-end AI platform. Both Amazon and eBay have been criticized in recent months over so-called “price-gouging” practices, whereby sellers vastly inflate the price of high-demand gadgets akin to face masks and hand-sanitizers. As issues transpired, the corporate’s current system was already working fashions of 100 million uploaded product photos at at time, and this proved helpful for every of eBay’s groups that have been seeking to automate the method of taking down listings that violated its insurance policies round price-gouging and prohibited gadgets.
“We created a really deep understanding of those images, where object detection, image classification, and textual recognition is available for the rest of the different business units,” Rawashdeh stated. “They didn’t need to worry about getting those expensive models — they were already available, and we leveraged them. As a result, we were able to take down over 15 million items on eBay.”