Real-world grounding in agentic AI
Amazon Science Homepageby Rose Yu8 Jun 2026
The year 2026 marks a definitive shift in the AI landscape: we have moved from models that simply know to agents that do.
Amazon Science Homepage
amazon.scienceadded August 2026
Bridging intent and execution in agentic systems
Amazon Science Homepageby Gaurav Gupta; Vatshank Chaturvedi8 Jun 2026
AI agent performance is not just a modeling problem; it is fundamentally a systems problem. A modern agent combines an LLM with a harness, software that mediates the LLM’s interaction with tools...
Ground truth is a process, not a dataset
Amazon Science Homepageby Venkatesh Saligrama3 Jun 2026
Today, the key challenge in AI isn’t only how to build better models; it’s how to build evaluation systems that can keep up.
How flat is replacing fat in AWS data center networks
Amazon Science Homepageby Giacomo Bernardi; Ratul Mahajan; Seshadhri Comandur28 May 2026
Routing in today’s data centers is usually governed by a data structure called a “fat tree”, which is similar to a corporate organizational chart, with nodes in each layer connecting to multiple...
Fall 2025 Amazon Research Awards recipients announced
Amazon Science Homepageby Amazon Research Awards team27 May 2026
Amazon Research Awards (ARA) provides unrestricted funds and AWS Promotional Credits to academic researchers investigating various research topics in multiple disciplines.
Diverse reasoning traces teach LLMs to make better decisions
Amazon Science Homepageby Sheng Jia; Xiao Wang; Shiva Kasiviswanathan26 May 2026
Large language models (LLMs) are pretrained on huge volumes of unlabeled data, but afterward, they’re typically post-trained on specific tasks such as instruction following, avoiding harmful...
Making LLMs faster without sacrificing accuracy
Amazon Science Homepageby Tao Yu; Youngsuk Park15 May 2026
Large language models (LLMs) keep getting bigger and better. But the cost of running them — generating text, answering questions, powering real-time applications — is scaling up, too.
Promptimus: Improving already good LLM prompts with zero manual engineering
Amazon Science Homepageby Zhengyuan Shen; Yunfei Bai; Sullam Jeoung; Shuai Wang14 May 2026
Large language models (LLMs) have become integral to enterprise applications across industries. Under the hood, customers’ inputs to the models are usually augmented with prompts that encode...
Navigating uncertainty in Amazon's middle-mile network
Amazon Science Homepageby Ruth Misener; Hana Ku; Georgios Paschos6 May 2026
Before the "last mile" delivery driver sets off for your home, your Amazon item has moved through the middle-mile network of fulfillment centers and sort centers, which brings products close enough...
How mechanism design theory helps optimize Amazon-vendor collaboration
Amazon Science Homepageby Dirk Bergemann5 May 2026
When Amazon places a purchase order with a vendor, a deceptively simple question arises: how many units should go to which fulfillment center, and when?
Building trust into AI
Amazon Science Homepageby Staff writer4 May 2026
At Amazon, AI now touches everything from warehouse logistics to customer service chatbots to AWS cloud services used by thousands of enterprises, making it a business-critical technology.
1 player call
top 17%plop score, higher is better
top 17%plop score, higher is betterPreserving the privacy of AI training data
Amazon Science Homepageby Parker Newton; Raj Copparapu29 Apr 2026
Large language models, the highest-profile machine learning (ML) models used today, are trained on huge corpora of public data.
How catastrophic is your LLM?
Amazon Science Homepageby Qian Hu; Weitong Ruan; Rahul Gupta27 Apr 2026
As large language models (LLMs) become increasingly useful across a variety of domains, the stakes of keeping them safe rise accordingly.