Amazon launches generative AI to help sellers write product descriptions
AI capabilities make it easier for sellers to write engaging, effective product listings and help shoppers find what they’re looking for.
Powering online shopping for global customers
Amazon Selection and Catalog Systems (ASCS) builds and operates the systems that host the world’s largest e-commerce product catalog. We power the online shopping experience, allowing customers to find, discover, and buy anything they want, anytime.
Our mission is to grow product selection and improve product data to provide world-class customer experience. We do this at scale by developing state-of-the-art systems, science, and processes.
What we do
Delivering at scale
Our systems and models process hundreds of millions of updates on billions of products. We’re at the cutting edge of the latest advances in artificial intelligence (AI) and machine learning (ML), pioneering large language model (LLM) and generative AI applications.
As Amazon grows and diversifies, the challenges get bigger, requiring us to continuously innovate. There are no textbook solutions at our scale.
ASCS by the numbers
- We manage tens of billions of products across hundreds of billions of SKUs.
- We run tens of thousands of ML models and push boundaries of ML science (e.g. generative AI and large language models, multi-modal, language agnostic processing).
- Our product taxonomy has hundreds of thousands of destination nodes with hundreds of billions of product-to-node assignments.
- We process hundreds of thousands of requests per second through our Tier-1 services with four nines availability and single digit millisecond latency (at P90).
Meet Amazonians working in ASCS
Rashmi
Rashmi started her journey at Amazon India in 2011. In 2016 she joined Items and Offers Platform (IOP) in California as a software development engineer (SDE). Since then, she’s progressed to software development manager. She currently manages the Catalog Data Management team within IOP.
Amir
Amir’s team mission is to develop scalable machine learning algorithms and create software solutions to enhance the shopping experience by monitoring and improving tens of billions of products across countries.
Saurabh
Saurabh joined Amazon in 2020 on the Items and Offers Platform (IOP) team.
Surbhi
Surbhi grew up in India in a very small town, where there were no proper schools for girls after eighth grade. She ended going to a boarding school at the age of 11 to pursue her love of math, which nobody understood in her hometown. She studied computer science in undergrad and grad school and moved to the US to work as a software developer.
Prithvi
Prithvi joined Selection Monitoring (SM) as a principal applied scientist in 2022. Previously, he was a researcher at IBM, where he published more than 50 research articles and 20 patents.
Learn about our team's innovations
How Amazon and NVIDIA help sellers create better product listings with AI
Amazon doubles inference speeds for new AI capabilities using NVIDIA TensorRT-LLM and GPUs to help sellers optimize product listings faster.
Amazon is encouraging sellers to use AI-generated product listings
ASCS is at the forefront of applying LLMs to enhance the Amazon Catalog.