AWS Machine Learning Blog

This post demonstrates a comprehensive observability solution using Amazon Managed Grafana dashboards that provides a holistic view of both quality and quantity for LLMs served on Amazon SageMaker AI endpoints with inference components.

Azercell Telecom LLC, Azerbaijan’s leading telecommunications provider, wanted to build an Azerbaijani large language model (LLM) on Amazon SageMaker AI for telecom use cases and a customer-facing chatbot. The challenge: adapting foundation models (FMs) to a morphologically rich language with limited training data and no existing blueprint for efficient LLM training in Azerbaijani. In a six-week collaboration, Azercell worked with the AWS Generative AI Innovation Center to establish a production-ready framework on Amazon SageMaker AI.

In this post, you learn how to build a custom portal with embedded SageMaker AI MLflow Apps UI. You walk through the architecture pattern behind a React front end paired with a Flask reverse proxy that handles AWS Signature Version 4 (SigV4) authentication, deploy the entire stack through the AWS Cloud Development Kit (AWS CDK), validate the deployment, and review security considerations and cleanup procedures.

In this post, we demonstrate how to build a secure Flask-based MLflow proxy service that provides HTTPS access to Amazon SageMaker MLflow without requiring the MLflow SDK. This solution is for organizations undergoing cloud transformation who want to preserve their existing ML workflows while adopting cloud-native services.

This post combines learnings from LangChain’s work on evaluating deep agents and Anthropic’s guide to demystifying evals for AI agents into a practical guide. In this post, you will learn how to: 1) apply five evaluation patterns for deep agents, 2) build offline evaluations using pytest and LangSmith, and 3) configure online monitoring for production. The walkthrough uses a text-to-SQL deep agent with Amazon Bedrock for the full development to production lifecycle.

Agent evaluation is most powerful when you combine fast-moving online signals with stable offline baselines. To understand whether your agent is truly improving over time, you need a fixed benchmark alongside your changing real-world traffic. Managing test cases for evaluation baselines as a dataset in Amazon Bedrock AgentCore brings the discipline of versioned test fixtures

This post covers Opus 4.8’s improvements and practical guidance for AI engineers integrating the model into agentic systems and production inference workloads on Amazon Bedrock.

This post demonstrates that integration in action by automating one of the most labor-intensive workflows in financial services: anti-money laundering (AML) alert triage. You will build a triage workflow using Amazon Quick Flows and Snowflake Cortex, connected through the Amazon Quick Model Context Protocol (MCP) integration. In our testing environment, automated workflows built using Amazon Quick reduced alert investigation time from 30-90 minutes to under 5 minutes. Actual results may vary based on alert complexity and data volume.

In this post, we explore how Amazon Bedrock Data Automation can accurately extract information from four common types of financial documents: bank statements, W-2 forms, 1099-B tax forms, and vendor contracts. We highlight the complexity in the documents, detail the custom extraction created in Amazon Bedrock Data Automation, and describe the outcomes of the extraction process.

In this post, we share how the AWS Generative AI Innovation Center (GenAIIC) collaborated with Works Human Intelligence (WHI) to build two AI agents using Amazon Bedrock AgentCore. We discuss the challenges encountered and the solutions that reduced costs by up to 97% while improving operational efficiency.

In this post, we show you how Verizon Connect built and scaled an agentic AI solution to transform overwhelming fleet data into clear, actionable insights for 100,000 users daily. We walk you through the architectural decisions, implementation challenges, and measurable results that can guide your own data-to-insights transformation.

In this post, we share how we built NarrateAI using Amazon Bedrock AgentCore to deliver business intelligence at scale for the AWS SMGS (Sales, Marketing and Global Services) organization. You will learn about: the two-layer architecture that separates batch processing from real-time interaction, the specialized AI agents that power intelligent routing and validation, key engineering patterns for production deployment, and how to build similar solutions with AWS services.

As agent adoption scaled, we saw a common pattern emerge across enterprises, including our own sales organization: specialized agents deliver value, but without orchestration, users carry the cognitive load of choosing between them. At AWS Sales, this meant more than 20 domain-specific agents deployed across the global organization, with representatives context-switching between systems instead of

Amazon Bedrock AgentCore payments is now available in preview, it provides instant payments to paid external services with no manual billing setup per provider, stablecoin support for cost-effective microtransactions that make sub-cent transactions economically viable, and configurable spending guardrails that give you fine-grained control over agent budgets and transaction limits. In this post, we walk you through a technical deep dive of AgentCore payments.

In this post, we provide a solution to build highly scalable, serverless multi-agent generative AI systems on AWS using LangGraph Agents as orchestrators integrated with Amazon Bedrock AgentCore Memory and Amazon Bedrock AgentCore Observability.

In this post you’ll learn how to build a multi-agent campaign review system that demonstrates parallel reasoning, context persistence, and traceable execution paths using an integrated architecture that combines NVIDIA NIM for GPU-accelerated inference. Amazon Bedrock AgentCore provides managed runtime, shared memory and built-in observability and Strands Agents provide serverless multi-agent orchestration. This approach supports performance, scalability, and operational insight in production environments. While the example focuses on marketing content review, the same pattern applies to digital assistants, review automation, and retrieval-augmented generation pipelines.

In this post, we demonstrate the capabilities of AgentWatch through practical implementation. You will see how the solution performs infrastructure checks every 15 minutes, summarizing CloudWatch metrics, logs, and alarms across multiple AWS accounts. The agent delivers actionable reports directly to Slack and responds to natural language queries about your infrastructure state. Throughout, we explore three human-in-the-loop patterns that maintain appropriate oversight while maximizing automation.

Building an AI app shouldn’t require a PhD in machine learning (ML) or months of wrestling with complex architectures. Yet that’s exactly what happens when you try to orchestrate multiple API calls, manage conversation state, and create agents that can reason on their own. I’ve seen straightforward AI ideas balloon into sprawling projects that demand

When hundreds to thousands of users are onboarded to an enterprise AI platform, business leaders and platform owners need visibility into who is using the platform, whether users are satisfied with the answers they receive, and which capabilities are driving the most engagement. Without a centralized observability solution, this data is scattered across multiple AWS

In this post, we explore how the Amazon Quick document and visualization creation capabilities work, what you can build with them, and how professionals across roles are using them to reclaim hours of their workweek. From technical execution to strategic judgment Most professional roles carry an unspoken assumption that a significant portion of your time