From AWS Machine Learning Blog · 19 stories
Implementing Defense-in-Depth Authorization for MCP Tools on Amazon Quick
A multi-gate authorization pattern for Model Context Protocol (MCP) tools on Amazon Quick has been implemented to provide defense-in-depth authorization. This pattern evaluates OpenID Connect (OIDC) JSON Web Token (JWT) claims sequentially to enforce role-based and attribute-based access control, addressing compliance needs beyond basic SSO authentication.
Automating User-Level Custom Permissions in Amazon Quick
Amazon Quick now allows automation of user-level custom permissions to enforce fine-grained access control. This enables organizations to maintain the principle of least privilege as their Quick environments scale and AI capabilities expand. The article details four architectural patterns for automating custom permission assignments across different user lifecycle stages.
Customizing Amazon Bedrock knowledge bases for complex documents using Amazon Textract
Amazon published a guide on integrating Amazon Textract with Amazon Bedrock to improve knowledge base accuracy for large and complex documents. This integration helps customer service teams extract information from varied document formats like utility bills, addressing issues like delayed responses and billing errors.
Embed Amazon QuickSight Visuals in React Apps with Cognito User Authentication
A new tutorial details how to embed individual Amazon QuickSight visuals into React applications using Amazon Cognito for per-user authentication. This method allows for fine-grained access control and granular layout management by integrating specific charts and graphs directly into an application interface.
Building a Restaurant Telephony AI Host with Amazon Connect
Amazon Web Services published a guide on how to build an AI-powered voice ordering system for restaurants using Amazon Connect. This system allows customers to place orders over the phone without human intervention, addressing a common challenge for restaurants with high call volumes.
Automating Document Processing with AWS IDP Accelerator and Amazon Quick Automate
AWS details how organizations can automate document processing pipelines using the AWS Generative AI Innovation Center (GAIIC) Intelligent Document Processing (IDP) Accelerator and Amazon Quick Automate. This approach aims to reduce manual handling and improve efficiency in document-intensive operations across various industries, using mortgage lending as a primary example.
AWS details KnowledgeForge, a solution for ITSM knowledge base management using generative AI
AWS released details on KnowledgeForge, a solution that uses generative AI to extract knowledge from resolved IT Service Management (ITSM) tickets and improve existing knowledge base articles. This solution helps organizations automate the creation of new knowledge base content and maintain the quality of their existing documentation.
Customizing Amazon Quick Embedded Chat for Application Integration
Amazon Quick embedded chat now offers customization options for visual theming and communication tone, allowing developers to integrate the conversational AI interface more seamlessly into their web applications. This update addresses the need for a consistent brand experience when embedding AI chat functionalities, preventing a disjointed user interface.
Jumio Implements Real-Time Feature Store on AWS for Fraud Detection
Jumio, an identity verification provider, built a real-time feature store on AWS to address challenges like data duplication, inconsistent feature definitions, and latency in its fraud detection machine learning models. This implementation uses Amazon SageMaker Feature Store, Amazon Managed Service for Apache Flink, and Amazon Kinesis Data Streams to provide sub-100ms latency for real-time predictions.
Designing Custom Reward Functions for Multi-Turn Reinforcement Learning with Amazon Nova Forge
Amazon Nova Forge now offers a generally available serverless multi-turn reinforcement learning (RL) option, alongside its Bring Your Own Orchestration (BYOO) capability, to help users define custom reward functions for multi-turn RL. This development provides more flexibility for training Amazon Nova models in complex, multi-step agentic tasks, addressing a key challenge in customizing these models.
Tutorial: Amazon Bedrock Cost Attribution with Athena and CUDOS
A tutorial details how to visualize and analyze Amazon Bedrock costs using Amazon Athena queries and CUDOS dashboards. This process involves setting up Cost and Usage Report (CUR) 2.0 with IAM principal data and then querying this data for granular cost attribution. This enables per-user and per-application visibility of Bedrock spend.
AWS details explainable next-best-product recommendation system for banking
AWS published an architectural overview for building an explainable next-best-product recommendation system for banking using Amazon SageMaker AI and PyTorch. This system helps banks predict customer product needs by analyzing transaction histories, product ownership, demographics, and behavioral patterns. The architecture uses a multi-tower neural network and learned attention for per-customer explainability, addressing challenges with traditional rule-based systems.
Building Multi-Region Visualizations with Highcharts in Amazon Quick Sight
This article details how to integrate Highcharts as a custom visual within Amazon Quick Sight to create multi-region data visualizations. This integration addresses limitations in native Quick Sight visualizations when analyzing complex, geographically diverse carrier performance data.
MRH Trowe Deploys Secure Self-Service AI Agents for 400 Employees in Financial Services
MRH Trowe, a German insurance broker, implemented a secure self-service AI agent system for approximately 400 employees using Strands Agents, Amazon Bedrock AgentCore, and LibreChat. This deployment addresses the need for AI tools that can handle sensitive client data within a governed, auditable, and cost-transparent environment in the financial sector.
Boomi Scribe Automates Integration Documentation Using AWS AI/ML Services
Boomi introduced Boomi Scribe, an AI-powered agent running on AWS that automates documentation for enterprise integration processes. This tool generates detailed documentation as developers work and provides version comparisons, addressing challenges with manual documentation and compliance.
AWS Named a Leader in Forrester Wave for AI Infrastructure Solutions Q4 2025
AWS has been recognized as a Leader in The Forrester Wave: AI Infrastructure Solutions, Q4 2025, receiving the highest score in the Strategy category. This evaluation assessed 13 providers on current offering, strategy, and customer feedback for AI infrastructure capabilities.
First Orion adopts Amazon Nova Act for AI-driven QA automation
First Orion implemented Amazon Nova Act to automate its quality assurance (QA) processes, shifting from script-based testing to AI-driven agents. This change addresses the challenge of QA keeping pace with rapid software development velocity, particularly as the company expanded its web applications.
Yahoo integrates Amazon Bedrock to enhance search retargeting in its DSP
Yahoo has integrated Amazon Bedrock into its Demand-Side Platform (DSP) to improve Search Retargeting (SRT) capabilities. This integration allows Yahoo to use generative AI for better understanding of search intent and more relevant ad targeting, moving beyond the limitations of its previous Word2Vec + LSH approach.
Guardoc Health uses Amazon Nova models to automate medical document processing
Guardoc Health is using Amazon Nova models, available through Amazon Bedrock, to automate the extraction, classification, and processing of complex medical documents for long-term care facilities. This implementation has led to a 46 percent reduction in documentation errors and a 70 percent decrease in audit fines for its clients. The technology addresses challenges with fragmented and inconsistent clinical documentation, aiming to improve patient outcomes and regulatory compliance.