Connect with us


How the pandemic readied Alibaba’s AI for the world’s biggest shopping day



The news: While the US has been hooked on its election, China has been shopping. From November 1 to 11, the country’s top e-commerce giants, Alibaba and JD, generated $115 billion in sales as part of their annual Single’s Day shopping bonanza. Alibaba, who started the festival in 2009, accounted for $74.1 billion of those sales, a 26% increase on last year. For comparison, Amazon’s 48-hour Prime Day sales only crossed the $10-billion mark this year.

Pandemic stress test: The sheer scale of the event makes it somewhat of a logistical miracle. To pull off the feat, Alibaba and JD invest heavily in AI models and other technology infrastructure to predict shopping demand, optimize the global distribution of goods across warehouses, and streamline worldwide delivery. The systems are usually tested and refined throughout the year before being stretched to their limits during the actual event. This year, however, both companies faced a complication: accounting for changes in shopping behavior due to the pandemic.

Broken models: In the initial weeks after the coronavirus outbreak, both companies saw their AI models behaving oddly. Because the pandemic struck during the Chinese New Year, hundreds of millions of people who would have otherwise been holiday shopping were instead buying lockdown necessities. The erratic behavior made it impossible to rely on historical data. “All of our forecasts were no longer accurate,” says Andrew Huang, general manager of the domestic supply chain at Cainiao, Alibaba’s logistics division.

People were also buying things for different reasons, which was flying in the face of the platforms’ product recommendations. For example, JD’s algorithm assumed people who bought masks were sick and so recommended medicine, when it might have made more sense to recommend them hand sanitizer.

Changing tack: The breakdown of their models forced both companies to get creative. Alibaba doubled down on its short-term forecasting strategy, says Huang. Rather than project shopping patterns based on season, for example, it refined its models to factor in more immediate variables like the previous week of sales leading up to major promotional events or external data like the number of covid cases in each province. As livestreaming e-commerce (showing off products in real time and answering questions from buyers) exploded in popularity during quarantine, the company also built a new forecasting model to project what happens when popular livestream influencers market different products.

And JD retooled its algorithms to consider more external and real-time data signals, like covid case loads, news articles, and public sentiment on social media.

Unexpected boon: Adding these new data sources into their models seems to have worked. Alibaba’s new livestreaming AI model, for example, ended up playing a core role in forecasting sales after the company made livestreaming a core part of its Single’s Day strategy. For JD, its updates may have also increased overall sales. The company says it saw a 3% increase in click-through rate on its product recommendations after it rolled out its improved algorithm, a pattern that held up during Single’s Day.

Understanding context: Both companies have learned from the experience. For example, Huang says his team learned that each livestream influencer mobilizes its fan base to exhibit different purchasing behaviors, so it will continue to create bespoke prediction models for each of its top influencers. Meanwhile, JD says it has realized how much news and current events influence e-commerce patterns and will continue to tweak its product recommendation algorithm accordingly.

Continue Reading


Teardown of “Dishy McFlatface,” the SpaceX Starlink user terminal



The outer part of

Enlarge / Ken Keiter gets ready to tear apart the SpaceX Starlink user terminal, “Dishy McFlatface.” (credit: Ken Keiter)

Engineer Ken Keiter recently came into possession of one SpaceX Starlink user terminal, the satellite dish that SpaceX nicknamed “Dishy McFlatface.” But instead of plugging it in and getting Internet access from SpaceX’s low Earth orbit (LEO) satellites, Keiter decided to take Dishy apart to see what’s inside.

The teardown process destroyed portions of the device. “I would love to actually test out the [Starlink] service and clearly I didn’t get a chance to, as this went a little bit further than I was intending,” Keiter said toward the end of the 55-minute teardown video he posted on YouTube last week.

Keiter, who lives in Portland, Oregon, was impressed by the Starlink team’s work. “It’s rare to see something of this complexity in a consumer product,” he said in reference to the device’s printed circuit board (PCB), which he measured at 19.75″ by 21.5″.

Read 10 remaining paragraphs | Comments

Continue Reading


Salesforce applies AI to workflow with Einstein Automate



While Salesforce made a big splash yesterday with the announcement that it’s buying Slack for $27.7 billion, it’s not the only thing going on for the CRM giant this week. In fact Dreamforce, the company’s customer extravaganza is also on the docket. While it is virtual this year, there are still product announcements aplenty and today the company announced Einstein Automate, a new AI-fueled set of workflow solutions.

Sarah Franklin, EVP & GM of Platform, Trailhead and AppExchange at Salesforce says that she is seeing companies facing a digital imperative to automate processes as things move ever more quickly online, being driven there even faster by the pandemic. “With Einstein Automate, everyone can change the speed of work and be more productive through intelligent workflow automation,” she said in a statement.

Brent Leary, principal analyst at CRM Essentials says that combined these tools are designed to help customers get to work more quickly. “It’s not only about identifying the insight, it’s about making it easier to leverage it at the the right time. And this should make it easier for users to do it without spending more time and effort,” Leary told TechCrunch.

Einstein is the commercial name given to Salesforce’s artificial intelligence platform that touches every aspect of the company’s product line, bringing automation to many tasks and making it easier to find the most valuable information on customers, which is often buried in an avalanche of data.

Einstein Automate encompasses several products designed to improve workflows inside organizations. For starters, the company has created Flow Orchestrator, a tool that uses a low-code, drag and drop approach for building workflows, but it doesn’t stop there. It also relies on AI to provide help suggest logical next steps to speed up workflow creation.

Salesforce is also bringing Mulesoft, the integration company it bought for $6.5 billion in 2018 into the mix. Instead of processes like a mortgage approval workflow, the Mulesoft piece lets IT build complex integrations between applications across the enterprise, and the Salesforce family of products more easily.

To make it easier to build these workflows, Salesforce is announcing the Einstein Automate collection page available in AppExchange, the company’s application marketplace. The collection includes over 700 pre-built connectors so customers can grab and go as they build these workflows, and finally it’s updating the OmniStudio, their platform for generating customer experiences. As Salesforce describes it, “Included in OmniStudio is a suite of resources and no-code tools, including pre-built guided experiences, templates and more, allowing users to deploy digital-first experiences like licensing and permit applications quickly and with ease. ”

Per usual with Salesforce Dreamforce announcements, the Flow Orchestrator being announced today won’t be available in beta until next summer. The Mulesoft component will be available in early 2021, but the OmniStudio updates and the Einstein connections collection are available today.

Continue Reading


Virta Health’s behavioral diabetes treatment service is now worth over $1 billion



A new $65 million investment led by the growth capital and public investment arm of Sequoia Capital will give Virta Health, a developer of a behavioral-focused diabetes treatment, a valuation of over $1 billion.

Virta’s approach, which uses a combination of approaches to change diet and exercise to reverse the presence of type 2 diabetes and other chronic metabolic conditions, has shown clinical success and attracted 100 health care payers to endorse the company’s treatments.

“We partnered with Virta for their ability to deliver unmatched health improvement and cost savings—two clear differentiators from other offerings on the market,” said William Ashmore, CEO of the State Employees’ Insurance Board of Alabama, in a statement. “Especially amid the COVID-19 pandemic, it’s vital that we provide our members the life-changing results Virta is known for delivering, through expert, virtual care delivered right to their home.”

The company said it would use the funding to expand sales and marketing efforts for its services as well as expand its research and development into other non-pharmaceutical therapies for metabolic conditions.

The financing came from Sequoia Capital Global Equities and Caffeinated Capital and brings the company’s total funding to over $230 million and gives it a $1.1 billion valuation, according to a statement.

Alongside Sequoia Capital Global Equities, Caffeinated Capital participated in the round, which brings total funding to more than $230 million and values Virta Health at over $1.1 billion.

Diabetes has long been an attractive condition for startups and has been the first target that companies focused on behavior changes to influence metabolic conditions aim to address. The reason why there are so many diabetes-focused businesses is because of the prevalence of the disease in the U.S. Almost half of adults in the U.S. suffer from obesity, pre-diabetes, or type 2 diabetes and the disease kills thirty people every hour. Diabetes also doubles the risk of death from COVID-19 infections.

Beyond the risks, the costs of treatment are skyrocketing. According to data from the American Diabetes Association released in March 2018, the total costs of treating diagnosed diabetes have risen to $327 billion in 2017 from $245 billion in 2012, when the cost was last examined.

“Given the scope of the metabolic crisis in the U.S. and globally, it cannot be understated how game-changing Virta’s results and care delivery are,” said Patrick Fu, managing partner at Sequoia Capital Global Equities, in a statement. “Virta’s technology-driven, non-pharmaceutical approach has fundamentally changed how diabetes is cared for, and our collective belief in what is possible for population health improvement. This is the future of chronic disease care.”

Continue Reading