The language AI agents might speak, sharing context without compromising privacy, modeling agentic negotiations, and understanding users’ commonsense policies are some of the open scientific questions that researchers in agentic AI will need to grapple with.

Amazon Science

03.09.2025

Simplifying book discovery with ML-powered visual autocomplete suggestions - Amazon Science

Audible's ML algorithms connect users directly to relevant titles, reducing the number of purchase steps for millions of daily users.

Amazon Science

27.08.2025

Revolutionizing warehouse automation with scientific simulation - Amazon Science

With a novel parallel-computing architecture, a CAD-to-USD pipeline, and the use of OpenUSD as ground truth, a new simulator can explore hundreds of sensor configurations in the time it takes to test just a few physical setups.

Amazon Science

25.08.2025

A decade of database innovation: The Amazon Aurora story - Amazon Science

From reimagining storage to serverless computing, Aurora continues to push the boundaries of what's possible in database technology.

Amazon Science

12.08.2025

Amazon builds first foundation model for multirobot coordination - Amazon Science

Trained on millions of hours of data from Amazon fulfillment centers and sortation centers, Amazon’s new DeepFleet models predict future traffic patterns for fleets of mobile robots.

Amazon Science

09.08.2025

A better path to pruning large language models - Amazon Science

A new philosophy for developing LLM architectures reduces energy requirements, speeds up runtime, and preserves pretrained-model performance.

Amazon Science

08.08.2025

Three challenges in machine-based reasoning - Amazon Science

Translating from natural to structured language, defining truth, and definitive reasoning remain topics of central concern in automated reasoning, but Amazon Web Services’ new Automated Reasoning checks help address all of them.

Amazon Science

01.08.2025

Multiagent AI for generating chain-of-thought training data - Amazon Science

Using ensembles of agents to generate and refine interactions annotated with chains of thought improves performance on a battery of benchmarks by an average of 29%.

Amazon Science

30.07.2025

Measuring the effectiveness of software development tools and practices - Amazon Science

New cost-to-serve-software metric that accounts for the full software development lifecycle helps determine which software development innovations provide quantifiable value.

Amazon Science

29.07.2025

Pushing the boundaries of secure AI: Winners of the Amazon Nova AI Challenge - Amazon Science

University teams battle to harden and hack AI coding assistants in head-to-head tournament

Amazon Science

29.07.2025

FalseReject: Reducing overcautiousness in LLMs through reasoning-aware safety evaluation - Amazon Science

Novel graph-based, adversarial, agentic method for generating training examples helps identify — and mitigate — "overrefusal".

Amazon Science

29.07.2025

Mitra: Mixed synthetic priors for enhancing tabular foundation models - Amazon Science

Generating diverse synthetic prior distributions leads to a tabular foundation model that outperforms task-specific baselines.

Amazon Science

29.07.2025

Pruning network nodes on the fly to improve LLM efficiency - Amazon Science

Language models inspired by specialized processing regions in the brain offer significant time and cost savings.

Amazon Science

16.07.2025

Unsupervised, generalizable method for doing anomaly detection - Amazon Science

An ensemble of models, weighted according to their reluctance to flag anomalies, outperforms its predecessors.

Amazon Science

26.06.2025

Using generative AI to do multimodal information retrieval - Amazon Science

With large datasets, directly generating data ID codes from query embeddings is much more efficient than performing pairwise comparisons between queries and candidate responses.

Amazon Science

25.06.2025

Finalist teams advance in the Amazon Nova AI Challenge: Trusted AI Track - Amazon Science

Top eight university teams move on to head-to-head finals focused on AI security for code generation.

Amazon Science

13.06.2025

Scaling up image segmentation across data and tasks - Amazon Science

Novel architecture that fuses learnable queries and conditional queries improves a segmentation model’s ability to transfer across tasks.

Amazon Science

04.06.2025

Amazon Research Awards recipients announced - Amazon Science

Awardees, who represent 44 universities in 10 countries, have access to Amazon public datasets, along with AWS AI/ML services and tools.

Amazon Science

30.05.2025

Independent evaluations demonstrate Nova Premier’s safety - Amazon Science

In both black-box stress testing and red-team exercises, Nova Premier comes out on top.

Amazon Science

22.05.2025

A first-of-its-kind experiment to measure the impact of out-of-home advertising - Amazon Science

By combining surveys with ads targeted to metro and commuter rail lines, Amazon researchers identify the fraction of residents of different neighborhoods exposed to the ads and measure ad effectiveness.

Amazon Science

13.05.2025

The path to better plastics: Our progress and partnerships - Amazon Science

How Amazon is helping transform plastics through innovation in materials, recycling technology, sortation, and more.

Amazon Science

10.05.2025

How Amazon’s Vulcan robots use touch to plan and execute motions - Amazon Science

Unique end-of-arm tools with three-dimensional force sensors and innovative control algorithms enable robotic arms to “pick” items from and “stow” items in fabric storage pods.

Amazon Science

03.05.2025

A better training method for reinforcement learning with human feedback - Amazon Science

Contrasting training pairs with large reward differences mitigate spurious correlations and improve performance of direct-alignment algorithms by as much as 20%–40%.

Amazon Science

02.05.2025

Your phone camera can autofocus. Why can't your specs? - Amazon Science

Startup IXI is working on a breakthrough in vision correction.

Amazon Science

02.05.2025

Amazon announces Ocelot quantum chip - Amazon Science

Prototype is the first realization of a scalable, hardware-efficient quantum computing architecture based on bosonic quantum error correction.

Amazon Science

30.04.2025

Video classifiers learn to recognize actions they've never seen - Amazon Science

New end-to-end approach to zero-shot video classification dramatically outperforms predecessors.

Amazon Science

12.04.2025

Automating hallucination detection with chain-of-thought reasoning - Amazon Science

Novel three-pronged approach combines claim-level evaluations, chain-of-thought reasoning, and classification of hallucination error types.

Amazon Science

11.04.2025

An AI agent for data science: Amazon Q Developer in SageMaker Canvas - Amazon Science

New tool lets customers build, train, and deploy machine learning models using only natural language.

Amazon Science

28.03.2025

Training large language models more efficiently - Amazon Science

Training separate models on different datasets and then merging them reduces computational costs by as much as 91%.

Amazon Science

22.03.2025

Empowering disaster preparedness: AI’s role in navigating complex climate risks - Amazon Science

AI systems that integrate meteorological, geospatial, and socioeconomic data can deliver warnings that are more localized and more timely.

Amazon Science

11.03.2025

Amazon Nova AI Challenge accelerating the field of generative AI - Amazon Science

Inaugural global university competition focused on advancing secure, trusted AI-assisted software development.

Amazon Science

01.03.2025

Using SAT solving to optimize quantum circuit mapping - Amazon Science

In experiments involving real quantum devices and algorithms, automated-reasoning-based method for mapping quantum computations onto quantum circuits is 26 times as fast as predecessors.

Amazon Science

21.02.2025

Training code generation models to debug their own outputs - Amazon Science

Using large language models to generate training data and updating models through both fine tuning and reinforcement learning improves the success rate of code generation by 39%.

Amazon Science

19.02.2025

Lightweight LLM for converting text to structured data - Amazon Science

Novel training procedure and decoding mechanism enable model to outperform much larger foundation model prompted to perform the same task.

Amazon Science

19.02.2025

Benchmarking tool for graph-centric predictive modeling on databases - Amazon Science

4DBInfer enables model comparison across datasets, predictive tasks, database-to-graph extraction methods, and graph-based predictive architectures.

Amazon Science

15.01.2025

QKD and authentication: Separating facts from myths - Amazon Science

Key exchange protocols and authentication mechanisms solve distinct problems and must be integrated in a secure communication system.

Amazon Science

25.12.2024

The 10 most viewed publications of 2024 - Amazon Science

From cloud databases and anomaly detection on graphs to recession prediction and Amazon's new Nova foundation models, these are the most viewed publications authored by Amazon scientists and collaborators in 2024.

Amazon Science

25.12.2024

The 10 most viewed blog posts of 2024 - Amazon Science

Large language models remained a hot topic, but posts about cryptography and automated reasoning also drew readers.

Amazon Science

21.12.2024

Amazon Research Awards recipients announced - Amazon Science

Awardees, who represent 10 universities, have access to Amazon public datasets, along with AWS AI/ML services and tools.

Amazon Science

20.12.2024

New AWS tool recommends removal of unused permissions - Amazon Science

IAM Access Analyzer feature uses automated reasoning to recommend policies that remove unused accesses, helping customers achieve “least privilege”.

home›Amazon Science›Software

Amazon Science - Software

Scientific frontiers of agentic AI - Amazon Science

Simplifying book discovery with ML-powered visual autocomplete suggestions - Amazon Science

Revolutionizing warehouse automation with scientific simulation - Amazon Science

A decade of database innovation: The Amazon Aurora story - Amazon Science

Amazon builds first foundation model for multirobot coordination - Amazon Science

A better path to pruning large language models - Amazon Science

Three challenges in machine-based reasoning - Amazon Science

Multiagent AI for generating chain-of-thought training data - Amazon Science

Measuring the effectiveness of software development tools and practices - Amazon Science

Pushing the boundaries of secure AI: Winners of the Amazon Nova AI Challenge - Amazon Science

FalseReject: Reducing overcautiousness in LLMs through reasoning-aware safety evaluation - Amazon Science

Mitra: Mixed synthetic priors for enhancing tabular foundation models - Amazon Science

Pruning network nodes on the fly to improve LLM efficiency - Amazon Science

Unsupervised, generalizable method for doing anomaly detection - Amazon Science

Using generative AI to do multimodal information retrieval - Amazon Science

Finalist teams advance in the Amazon Nova AI Challenge: Trusted AI Track - Amazon Science

Scaling up image segmentation across data and tasks - Amazon Science

Amazon Research Awards recipients announced - Amazon Science

Independent evaluations demonstrate Nova Premier’s safety - Amazon Science

A first-of-its-kind experiment to measure the impact of out-of-home advertising - Amazon Science

The path to better plastics: Our progress and partnerships - Amazon Science

How Amazon’s Vulcan robots use touch to plan and execute motions - Amazon Science

A better training method for reinforcement learning with human feedback - Amazon Science

Your phone camera can autofocus. Why can't your specs? - Amazon Science

Amazon announces Ocelot quantum chip - Amazon Science

Video classifiers learn to recognize actions they've never seen - Amazon Science

Automating hallucination detection with chain-of-thought reasoning - Amazon Science

An AI agent for data science: Amazon Q Developer in SageMaker Canvas - Amazon Science

Training large language models more efficiently - Amazon Science

Empowering disaster preparedness: AI’s role in navigating complex climate risks - Amazon Science

Amazon Nova AI Challenge accelerating the field of generative AI - Amazon Science

Using SAT solving to optimize quantum circuit mapping - Amazon Science

Training code generation models to debug their own outputs - Amazon Science

Lightweight LLM for converting text to structured data - Amazon Science

Benchmarking tool for graph-centric predictive modeling on databases - Amazon Science

QKD and authentication: Separating facts from myths - Amazon Science

The 10 most viewed publications of 2024 - Amazon Science

The 10 most viewed blog posts of 2024 - Amazon Science

Amazon Research Awards recipients announced - Amazon Science

New AWS tool recommends removal of unused permissions - Amazon Science

home
›
Amazon Science
›
Software