Automated AI newsroom

AI News

High-signal AI developments, independently scored, summarized and organized for fast reading.

Live intelligence

Editor-selected AI developments

19 updates

Tutorial Editor selected

Use decision models with Amazon Bedrock AgentCore Evaluations

This post explains how to integrate decision models like Jev and Strands Decider 2B into Amazon Bedrock AgentCore Evaluations via code-based evaluators. It covers the trade-offs between decision models and LLM-as-a-judge, provides a step-by-step validation procedure, and recommends a cascade pattern where decision models screen turns and escalate low-confidence cases to built-in evaluators. The accompanying code sample demonstrates an HR Assistant agent with deterministic, built-in, and decision-model evaluators.

Why it matters: Provides a concrete, recent AWS tutorial with practical guidance on using decision models for agent evaluation, including validation steps and architecture, valuable for developers building AI evaluation pipelines.

AWS Machine Learning Blog2d agoScore 85

AI evaluationagent evaluationdecision models

Tutorial Editor selected

Four layers of context engineering for production AI agents: How Deep Insight applies them in working code

This AWS blog post details a four-layer approach to context engineering for production AI agents, using the open-source Deep Insight multi-agent system as a reference. The layers are architecture-level context isolation, prompt-based output control, file-based context externalization, and verification with independent agents. Concrete code examples show a 95% token reduction in script context and a 97% reduction in system prompt size.

Why it matters: Provides actionable, code-backed techniques for managing LLM context windows in production, directly addressing a common bottleneck with measurable improvements.

AWS Machine Learning Blog2d agoScore 85

context engineeringmulti-agent architectureLLM context window

Tutorial Editor selected

Amazon Quick unifies structured and unstructured data querying

Amazon Quick now lets users connect S3-based unstructured documents and enriched structured datasets in a single Space, then query both through a custom chat agent. The walkthrough shows creating an S3 integration, building a knowledge base, enriching a CSV dataset with semantic context, and configuring an agent that answers cross-data questions with citations. It also covers generating artifacts like PowerPoints and sharing the agent.

Why it matters: Provides a concrete, recent tutorial on a new AWS feature that combines structured and unstructured data querying, valuable for data professionals.

AWS Machine Learning Blog3d agoScore 85

Amazon Quickunified data queryingknowledge bases

Tutorial Editor selected

ServiceNow on-behalf-of token exchange with Bedrock AgentCore Gateway

This post demonstrates how to configure Amazon Bedrock AgentCore Gateway for on-behalf-of (OBO) token exchange with Microsoft Entra ID, enabling AI agents to call ServiceNow REST APIs while preserving the user's identity. It covers setting up Entra application registrations, ServiceNow inbound OAuth integration, and AgentCore Gateway configuration with native OBO support. The pattern eliminates custom Lambda interceptors and applies to other OAuth-protected REST APIs.

Why it matters: Provides a concrete, recent technical guide for implementing identity propagation in AI agents, a critical enterprise requirement, with step-by-step configuration details.

AWS Machine Learning Blog3d agoScore 85

Amazon Bedrock AgentCoreOAuth 2.0AI agents

Tutorial Editor selected

Build a semantic layer for agentic AI with EDB Postgres AI and Amazon Bedrock AgentCore

This post demonstrates how to build a semantic layer for agentic AI using EDB Postgres AI, where business meaning is stored inside the operational Postgres database via comments, views, and semantic aliases. The layer is exposed to an external agent through an MCP endpoint, with governance enforced by Postgres roles, column-level grants, and row-level security. The architecture integrates with Amazon Bedrock AgentCore, and the post shows how different roles get different answers to the same question.

Why it matters: Provides a concrete, recent architecture for embedding a semantic layer in a database, with practical governance and observability details, valuable for AI engineers building agentic systems.

AWS Machine Learning Blog3d agoScore 92

semantic layeragentic AIdatabase governance

Tutorial Editor selected

Prepare web content for Amazon Bedrock Knowledge Bases with Crawl4AI

This post presents a solution for ingesting web content into Amazon Bedrock Knowledge Bases using Crawl4AI, an open-source Python crawler, when the built-in Web Crawler connector lacks control over content extraction. The workflow crawls public websites, converts pages to Markdown with metadata sidecars, uploads to S3, and triggers ingestion into a managed or customer-managed knowledge base. It includes a CDK deployment guide and validation steps for retrieval quality.

Why it matters: Provides a concrete, actionable pattern for custom web ingestion into Bedrock Knowledge Bases, addressing a common RAG pain point with a sample implementation.

AWS Machine Learning Blog4d agoScore 85

Amazon Bedrock Knowledge BasesWeb crawlingRAG data ingestion

Tutorial Editor selected

Build efficient agents with knowledge distillation on Amazon SageMaker AI

This post demonstrates how to use sequence-level knowledge distillation (SeqKD) to fine-tune a smaller student model (Qwen3.5-4B) to replace a frontier teacher (GLM-5.3) for a shopping agent, reducing cost and latency. It covers generating high-quality trajectories via rejection sampling, encoding them for supervised fine-tuning, running serverless training on SageMaker AI, and evaluating the student with a low-cost pre-check before full end-to-end evaluation. The approach is practical for API-only teachers and includes code snippets for the entire pipeline.

Why it matters: Provides a concrete, actionable guide for a cost-saving AI technique (SeqKD) with a real-world agentic use case, including implementation details on AWS.

AWS Machine Learning Blog4d agoScore 92

knowledge distillationagentic AImodel efficiency

Tutorial Editor selected

AWS builds multi-agent engineering assistant with Nova Multimodal Embeddings and Bedrock AgentCore

AWS presents a multi-agent system for aerospace part substitution evaluation, combining Amazon Nova Multimodal Embeddings, Amazon Bedrock AgentCore, and the Strands Agents SDK. The system routes queries across three specialist agents, uses deterministic code for spec-fit scoring, and keeps engineers as final decision-makers. It supports multimodal retrieval and visual similarity scoring, with a human-in-the-loop design for audit trails.

Why it matters: This is a concrete, recent technical guide with architecture details and code, valuable for engineers building multimodal RAG and multi-agent systems.

AWS Machine Learning Blog4d agoScore 85

Amazon BedrockMultimodal AIAgentic AI

Tutorial Editor selected

AWS publishes end-to-end Amazon Bedrock cost management playbook

AWS outlines a five-stage operating model for managing Amazon Bedrock generative AI costs: attribute spend to IAM principals, log requests with metadata, enforce budgets, optimize platform defaults, and run a review cadence. The playbook covers per-user attribution behind shared gateways, invocation logging with CloudWatch and S3/Athena, and three enforcement tiers ranging from alerts to synchronous gateway blocking. It also details model routing and prompt caching as central cost levers, with diagnostic queries for token usage by caller and workload.

Why it matters: It gives FinOps and platform teams a concrete, implementation-oriented framework for attributing, observing, and controlling Bedrock spend as agentic workloads increase token consumption.

AWS Machine Learning Blog6d agoScore 80

Amazon BedrockFinOpsAI cost governance

Tutorial Editor selected

Building a Governed Agentic AI Platform on Amazon Bedrock AgentCore

This post presents an enterprise architecture for an Agentic AI Platform built on Amazon Bedrock AgentCore, addressing the operational challenges of scaling agents across teams. It details a federated operating model with a Platform control plane and repeatable Workstream cells, covering governed model and tool access, delegated delivery, fleet operations, and lifecycle management. The blueprint includes multi-account topology, service control policies, CI/CD pipelines with evaluation gates, and defense-in-depth security controls.

Why it matters: Provides a concrete, actionable reference architecture for enterprises seeking to operationalize agentic AI at scale, addressing real governance and delivery bottlenecks.

AWS Machine Learning BlogOct 29, 2026Score 85

Amazon BedrockAgentic AIEnterprise Architecture

Tutorial Editor selected

Fixing evaluation bias for model selection on Amazon Bedrock

AWS and Property Finder describe how 96.8% of evaluation labels had been inherited from the incumbent model, biasing comparisons against alternatives. They built expert-annotated ground truth and used iterative prompt optimization to select Llama 4 Maverick 17B Instruct on Amazon Bedrock. The post reports development-set Micro F1 rising from 0.55 to 0.81 and an estimated 86% cost reduction, with a caveat that evaluation-set scores are optimistic because prompts were tuned on included images.

Why it matters: It gives practitioners a concrete methodology for auditing biased evaluation labels and running fairer model selection, with reported accuracy and cost tradeoffs.

AWS Machine Learning BlogOct 28, 2026Score 88

Amazon Bedrockmodel evaluationprompt optimization

Tutorial Editor selected

AWS details multi-turn reinforcement learning on SageMaker AI

AWS published a hands-on guide to Multi-Turn Reinforcement Learning (MTRL) on Amazon SageMaker AI, which trains models on end-of-episode rewards rather than per-step labels. Agents hosted on Bedrock AgentCore (or EKS, EC2, Fargate) play multi-step tasks and return a single reward, with GRPO updating a LoRA adapter. In the guide's Wordle example, gpt-oss-20b improved its solve rate on held-out words from 61% to 84% at default reasoning effort and from 15% to 69% at low effort, at roughly $339 in

Why it matters: It gives practitioners a concrete, cost-transparent walkthrough of agentic multi-turn RL on a major cloud platform, including benchmark numbers, supported models and failure modes.

AWS Machine Learning BlogOct 24, 2026Score 83

AWSAmazon SageMaker AIMulti-Turn Reinforcement Learning

Tutorial Editor selected

Measuring, not guessing, SageMaker autoscaling targets for TTS endpoints

This AWS Machine Learning Blog post walks through deploying the Kokoro-82M text-to-speech model as a custom container on an Amazon SageMaker AI real-time endpoint, then deriving autoscaling thresholds from two live measurements instead of guesswork. It measures Q_max (the per-instance concurrency ceiling that still meets a 3,000 ms p95 time-to-first-audio SLO, found to be 50 in the sample run) and T_total (end-to-end scale-out time, 260.2 s measured). The tooling then computes scale-out and scal

Why it matters: It gives practitioners a concrete, measurement-driven method and tooling for tuning autoscaling on latency-sensitive TTS inference endpoints, a gap that default LLM- and request/response-oriented guidance does not cover.

AWS Machine Learning BlogOct 23, 2026Score 80

Amazon SageMaker AIKokoro-82Mtext-to-speech

Tutorial Editor selected

AWS details production architecture for its open-source Deep Insight multi-agent system

AWS Solutions Architecture published the first part of a series on Deep Insight, an open-source multi-agent system that turns CSV data and natural-language prompts into validated analysis reports. The post describes eight role-specialized agents (Coordinator, Planner, Supervisor, Coder, Validator, Reporter, Tracker, Plan Reviewer) built with the Strands Agents SDK Graph and Agents-as-Tools patterns, with model selection tiered across Claude Haiku, Opus and Sonnet via Amazon Bedrock. It also outl

Why it matters: It offers concrete, reusable architectural guidance and trade-offs for teams moving multi-agent systems from prototype to production, backed by open-source code.

AWS Machine Learning BlogOct 22, 2026Score 80

AWSmulti-agent systemsagent orchestration

Tutorial Editor selected

AWS shows SFT-then-GRPO recipe to make GPT-OSS-20B reason in non-English languages

AWS details a fine-tuning workflow on Amazon SageMaker HyperPod that teaches GPT-OSS-20B to reason and answer in French, German, Spanish and Italian instead of defaulting to English. Supervised fine-tuning with LoRA raised reasoning-language accuracy from 22% to 100% on a 50-case evaluation, while adding GRPO on top of the SFT checkpoint lifted answer-language accuracy from 62% to 70%. A controlled Qwen2.5-7B experiment across multilingual reasoning and GSM8K tasks supports the argument that GRP

Why it matters: It offers a concrete, reproducible fine-tuning methodology and decision framework for choosing between SFT and GRPO when adapting open-weight reasoning models to non-English languages.

AWS Machine Learning BlogOct 21, 2026Score 83

Amazon SageMaker HyperPodGPT-OSS-20Breinforcement learning

Tutorial Editor selected

Push-based KV transfer in vLLM cuts TTFT up to 3x for disaggregated inference

AWS describes a push-based NIXL KV connector for vLLM's prefill-decode disaggregated serving, contributed upstream as an accelerator-agnostic feature. Instead of the decode instance pulling KV blocks after prefill, the decode side registers target blocks in advance and the prefill side issues an NIXL WRITE from a background thread, removing the router round trip and setup work from the TTFT critical path. On a 1P1D deployment with EC2 P5en instances and EFA, the Qwen3-32B benchmark showed up to

Why it matters: It gives concrete benchmark data and implementation guidance for a real vLLM latency optimization that teams running disaggregated inference can adopt.

AWS Machine Learning BlogOct 20, 2026Score 82

vLLMLLM inference servingAWS

Tutorial Editor selected

Bi-directional KV transfer in vLLM cuts TTFT for multi-turn agentic workloads

AWS details a bi-directional KV cache transfer capability in vLLM that lets decode nodes return retained KV blocks to prefill nodes on later conversation turns, avoiding repeated prefill computation in P-D disaggregated serving. It uses an extended NIXL KV connector and a KV-cache-aware router that tracks block metadata across turns, with configurable block retention TTL and a recompute threshold. On a 1P1D Amazon EC2 P5en deployment with EFA, AWS reports TTFT reductions of up to 1.6x for Qwen3-

Why it matters: Gives LLM serving teams a concrete open-source vLLM optimization, with configuration guidance and measured latency data, for reducing time to first token in long multi-turn agent sessions.

AWS Machine Learning BlogOct 20, 2026Score 84

vLLMKV cache transferLLM inference serving

Tutorial Editor selected

How ShipStation Global built an AI IT help desk on Amazon Bedrock AgentCore

ShipStation Global built Deckhand, a Slack-based AI IT help desk agent on Amazon Bedrock AgentCore, using Strands Agents, AgentCore Gateway MCP tools, and Amazon Bedrock Knowledge Bases. The system answers routine IT and company questions, reads error-message screenshots, and creates or updates Jira Service Management tickets, with logic for staying silent when humans are already handling a thread. In its first three months after an April 2026 launch, it handled 1,935 messages, replied to 97.5%

Why it matters: Offers a detailed, metrics-backed architecture for building a production agentic IT help desk on AWS.

AWS Machine Learning BlogOct 13, 2026Score 82

Amazon Bedrock AgentCoreAI agentsIT help desk automation