reMARS revisited: Amazon’s supply chain optimization
How the SCOT team implemented a system that leverages operations research and machine learning to decide what products to buy, how much to buy, where to place them, and more.
[Editor's note: INFORMS, the world's largest operations research and analytics conference, is occurring this week, so we're sharing this interview and presentation on how Amazon's Supply Chain Optimization Technologies team deployed one of the largest algorithmic decision-making machines in the world for supply chain optimization.]
In June 2022, Amazon re:MARS, the company’s in-person event that explores advancements and practical applications within machine learning, automation, robotics, and space (MARS), took place in Las Vegas. The event brought together thought leaders and technical experts building the future of artificial intelligence and machine learning, and included keynote talks, innovation spotlights, and a series of breakout-session talks.
Now, in our re:MARS revisited series, Amazon Science is taking a look back at some of the keynotes, and breakout session talks from the conference. We've asked presenters three questions about their talks, and provide the full video of their presentation.
On June 27, Deepak Bhatia, vice president of Supply Chain Optimization Technologies (SCOT), and Salal Humair, SCOT vice president and Amazon distinguished scientist, presented their talk, "Hands-off-the-wheel automation: Amazon’s supply chain optimization". Their session focused on a large-scale, hands-off-the-wheel automated system that leverages advanced mathematical optimization and machine learning to decide what products to buy, how much to buy, where to place them, and more.
What was the central theme of your presentation?
We shared the story of how Amazon continuously innovates to deliver better results to our customers, highlighted by the evolution of the SCOT team. We discussed how SCOT deployed one of the largest algorithmic decision-making machines in the world for supply chain optimization, how it helped Amazon scale worldwide, and how — while running the business — we went back to the drawing board to develop new technologies to both meet increasingly higher customer expectations, and to navigate an increasingly complex network.
This innovation is instrumental to giving customers access to the best selection while also delivering products to customers within two to 24 hours in a network that has more than doubled in size in the past five years.
In what applications do you expect this work to have the biggest impact?
SCOT systems influence customers’ experiences with the Amazon Store every day. We build our systems to directly impact these experiences — e.g., promise speed, delivery accuracy, in-stock rates, new selection — and seller experiences, and we relentlessly monitor that impact. We continuously seek to build new capabilities that increase satisfaction and drive the Amazon flywheel.
What are the key points you hope audiences take away from your talk?
Amazon and SCOT are pushing the frontier of science, tech, and product management for e-commerce supply chains. It is possible to run a business at our scale through technology and still innovate as if you are a startup — SCOT’s history is a prime example of this. While a lot has been accomplished, the field for innovation, and for impact, in teams like SCOT is still wide open.