Machine learning drives our algorithms for demand forecasting, product search ranking, product and deals recommendations, merchandising…

Judged as of April 2017. True if Amazon's disclosures or credible reporting show that by then Amazon used machine learning in all or nearly all of the named areas: demand forecasting, search ranking, product and deals recommendations, merchandising placements, fraud detection and translation. Mostly true if machine learning was used in most of these areas and one area is unconfirmed or was only at an early stage. Mixed if use is confirmed in only about half of the areas. Mostly false if use is confirmed in only one or two areas. False if evidence shows Amazon did not use machine learning in these functions.

Set before the research began.

1. E1 AI to fuel Amazon's growth: Bezos TechCentral (reporting Bloomberg) · 13 April 2017

   “Machine learning drives our algorithms for demand forecasting, product search ranking, product and deals recommendations, merchandising placements, fraud detection, translations and much more.”

2. E2 The evolution of Amazon's inventory planning system Amazon Science · 1 October 2021 Own record

   “Forecasting models developed by SCOT predict the demand for every product.”

3. E3 The evolution of Amazon's inventory planning system Amazon Science · 1 October 2021 Own record

   “Customer demand for products cannot be perfectly predicted even with the most advanced machine learning models.”

4. E4 Amazon blazes a trail to the next frontier in AI: the cloud - SiliconANGLE SiliconANGLE · 22 April 2017

   “Other machine learning work is behind the scenes, powering demand forecasting, product recommendations and more, and that’s where Bezos expects it to have the most impact.”

5. E5 Bringing the Magic of Amazon AI and Alexa to Apps on AWS. All Things Distributed (Werner Vogels blog) · 30 November 2016 Own record

   “Today, there are thousands of machine learning scientists and developers applying machine learning in various places, from recommendations to fraud detection, from inventory levels to book classification to abusive review detection.”

6. E6 Bringing the Magic of Amazon AI and Alexa to Apps on AWS. All Things Distributed (Werner Vogels blog) · 30 November 2016 Own record

   “There are many more application areas where we use ML extensively: search, autonomous drones, robotics in fulfillment centers, text processing and speech recognition (such as in Alexa) etc.”

7. E7 PowerPoint Presentation Amazon Web Services (AWS Summit Atlanta session EXC102) Own record

   “Circa 2007 Today Decision tree algorithm-driven model Convoluted neural networks”

The independent reports (E1, E4) only repeat the wording of Bezos's shareholder letter and do not check how any of the systems worked. The other support is all from Amazon. Later Amazon Science articles (E2, E3) describe demand-forecasting models, and an undated AWS slide deck (E7) contrasts a circa-2007 decision-tree model with neural networks today. These are Amazon's own ordinary statements, not independent checks or official records. Apart from a report repeating the shareholder letter's wording (E1), none of the evidence covers merchandising placements or translation.