LearnQuest

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).* In this advanced course, you explore the operational practices that keep containerized workloads reliable and performant long after the initial deployment. Working with both Amazon ECS and Amazon EKS, you implement progressive delivery strategies to reduce release risk, configure coordinated autoscaling across application and infrastructure layers, and build incident-ready observability using structured metrics, traces, and alerts. You develop failure-mode runbooks that prepare your team for real-world outages, then validate your skills through a final challenge scenario. By the end, you will operate container platforms with the confidence and rigor production environments demand.

This course provides the foundation learning for operational responsibility. In this course, you'll learn how AWS Elastic Beanstalk takes on and retains the operational responsibility for a web application. These responsibilities include provisioning, deployment, scaling, patching, monitoring, and recovery so that a small team can operate a portfolio of applications without a dedicated operations team.

Discover how to move beyond traditional 'vibe coding' by creating structured specifications, automating repetitive tasks, and orchestrating autonomous agents across complex codebases. In this course, you'll learn advanced features including Kiro Powers, multi-modal chat, and repository-level configurations for enterprise-grade development. You'll utilize hands-on demonstrations, explore steering files, agent hooks for workflow automation, and extend Kiro using Model Context Protocol (MCP) integration.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/aws-training).* Learn how to take an idea from concept to production with Kiro, an agentic IDE that helps you plan, build, and ship software through spec-driven development. This course walks through Kiro's core workflow: turning a rough idea into structured requirements, design, and tasks, using steering and hooks to encode team conventions, and verifying and shipping the result to AWS.

This lab teaches you about Amazon DynamoDB by walking you through how to create, query, view and delete a table in the AWS Management Console.

In this intermediate-level course, you explore how to build scalable applications using Amazon Keyspaces (for Apache Cassandra), a fully managed database service compatible with Apache Cassandra. You learn how to create keyspaces and tables, load data, and run queries using familiar Cassandra Query Language (CQL) tools. Through hands-on exercises, you gain practical experience designing and implementing NoSQL database solutions that leverage the serverless architecture of Amazon Keyspaces, helping you reduce operational overhead while maintaining high availability and performance.

In this course you learn about how Amazon Quick now delivers a unified experience across Desktop and Mobile, so your work follows you wherever you go. Start a conversation at your desk, pick it up on your phone — your conversations, feed, artifacts, and custom agents stay in sync seamlessly. No more lost context, no more starting over.

In this intermediate-level course, you explore how to build scalable, decoupled applications using Amazon Simple Queue Service (Amazon SQS). You learn how to create and configure standard and FIFO queues, send and receive messages, and implement common messaging patterns that improve application reliability. Through practical examples and hands-on guidance, you gain the skills to integrate Amazon SQS into your architectures, manage message lifecycle, and apply best practices for error handling and dead-letter queues. By the end of this course, you can confidently use Amazon SQS to decouple and scale your distributed systems.

This foundational course introduces business users to Mission control, a centralized hub in Amazon Quick where you can monitor your completed runs across chat agents, flows, apps, and automation executions, track activity across your shared assets, set up triggers to automate workflows, and manage tasks that require your attention. Learn how to use the core capabilities to streamline your workflow management and stay on top of all your Quick assets from one interface.

**What is a Lab Tutorial?** Lab Tutorial is a new BETA offering where you will use a personal AWS account to implement a solution. If you don't have an AWS account, you can watch the walkthrough videos to learn the lab steps. **Note:** If the lab has associated infrastructure costs, you are responsible for that cost of the AWS services’ usage while running this lab in your account. There is no cost associated with this lab tutorial itself. Once you complete the lab on your account, remember to clean-up the resources provisioned by following the clean-up instructions provided. In this lab tutorial, you will build a complete Model Context Protocol (MCP) server that extends Anthropic Claude's capabilities with custom tools, resources, and prompts. You learn how MCP revolutionizes AI integrations by shifting the burden of tool implementation from your application to dedicated servers, enabling powerful document management capabilities with minimal code. Throughout, you create server with tools to read and update documents, enabling you to chat with Claude about these documents in real-time.

This course teaches you how to integrate Palo Alto Networks Advanced DNS Security into AWS Route 53 Resolver DNS Firewall. You will learn to subscribe, configure partner managed rules across nine threat categories, monitor rule states, apply best practices for staged deployment, and manage the full subscription lifecycle — all without deploying additional infrastructure.

In this course, you'll learn how to choose and optimize the right AWS hardware for each stage of the AI pipeline, from data processing through training, inference, and application deployment. You'll explore the full range of AWS compute options for AI workloads, including AWS Graviton, AWS Trainium and Inferentia, GPU-based instances, Intel-, and AMD-based instances and learn when to use each to achieve optimal cost and performance.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/aws-training).* Amazon SageMaker AI is the AWS service for training, deploying, and orchestrating machine learning workflows at scale. This course teaches the SageMaker AI Python SDK v3, the current SDK (stable since May 2026), from first principles. v3 introduces a unified programming model that applies the same code shape to both traditional ML on tabular data and foundation model fine-tuning on open-weight LLMs. You will learn how to train models with ModelTrainer, deploy them with ModelBuilder, orchestrate multi-step workflows with SageMaker Pipelines, and fine-tune foundation models using the new SFTTrainer and DPOTrainer. One demo persona, AnyCompany Customer Insights, runs through every module, so you see the same SDK surface applied to two complementary ML solutions in the same business: a tabular escalation predictor built with XGBoost and a support triage agent built by fine-tuning Qwen3 1.7B for tool calling. For practitioners with existing SageMaker Python SDK v2 code, a dedicated migration module (Module 6) covers what changed and how to port forward. Newcomers can skip that module and cover the rest of the course as a self-contained SDK primer.

***This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).*** Deploy and compare neural audio codecs on Amazon SageMaker AI. Neural audio codecs convert speech into discrete tokens, the tokenization layer that speech language models are built on. In this lab you host two open codecs, Mimi and EnCodec, behind a single SageMaker AI real-time endpoint and benchmark them under identical conditions on frame rate, bitrate, and latency. You also explore the bitrate-quality trade-off by reconstructing the same clip at 1, 2, 4, and 8 codebooks, and learn when to choose real-time versus asynchronous inference. By the end you can deploy the codec layer yourself and size it to your own speech workload.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).* This delivers production-grade architecture patterns for demanding advertising workloads—covering core design principles, ad insertion (live/VOD/FAST), publisher ad server pipelines mapped to Well-Architected pillars, RTB architectures (Real-Time Bidder on EKS, PrivateLink, RTB Fabric), data pipelines from ingestion to reporting, and multi-region deployments addressing availability, DR, latency, and data residency. It also covers privacy-enhanced collaboration (Clean Rooms, TEEs), Customer Data Platform architecture (identity resolution, segmentation, activation), and Agentic GenAI use cases including ad creative generation, contextual advertising, adbreak detection, and intelligent ad operations powered by Amazon Bedrock.

This lab teaches you to deploy an agent to Amazon Bedrock AgentCore Runtime with AgentCore Identity inbound auth.

This assessment validates your comprehension of the Cloud Game Development Knowledge Badge Readiness Path. The assessment questions are based on the courses in this learning path. Take all of the courses and then verify your knowledge. Already have some knowledge on Cloud Game Development ? Go directly to the assessment, test your knowledge. You can earn the Cloud Game Development Knowledge badge with a score of 80% or better. Badges are issued by Credly within 5-7 business days. Don't worry if you didn't achieve 80% or better. You can re-take the assessment. Use this assessment as a method to continue building your knowledge. Assessment questions are randomized on each attempt which allows you to get more questions.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).* Learners will discover how the open source Amazon Bedrock Model Profiler simplifies the process of selecting foundation models by aggregating metadata -- including pricing, regional availability, context windows, and throughput -- into a single searchable interface. The course covers real-world scenarios the tool supports, such as optimizing cost, comparing model capabilities side by side, and accelerating production decisions. Participants also learn how to deploy the serverless web application in their own environment in under five minutes, reducing the manual effort of searching scattered documentation and console pages.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).* This self-paced course prepares AWS Certification Subject Matter Experts to contribute high-quality exam items during Item Development Workshops. Learners will develop the skills to write items from scratch and edit and review AI-generated items, applying AWS Certification standards for structure, quality, cognitive complexity, and fairness. The course covers item components, writing procedures, review processes, and submission workflows in ESO.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).* Learners will gain practical skills for tracking AI coding agent usage -- specifically Claude Code -- by shipping OpenTelemetry metrics directly to Amazon CloudWatch using bearer token authentication. The course covers end-to-end setup requiring no collectors or sidecars, enabling per-developer cost attribution, team-level usage analytics, and operational alerting queryable with PromQL. Participants will understand how bearer tokens work for tools running outside AWS, learn security considerations for long-term credentials, and discover how to answer critical questions about token consumption and cost that existing dashboards cannot address.

A closing encouragement to put your new skills to work, using AI tools available to you today in your everyday tasks.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).* Learners can gain practical skills in implementing object detection using Amazon Nova 2 Lite through Amazon Bedrock -- without the need for model training, large datasets, or dedicated data science teams. The course covers prompt engineering for object detection, processing structured JSON output with bounding box coordinates, and deploying a serverless application using AWS Lambda and Amazon API Gateway. Learners also explore real-world applications in manufacturing, agriculture, and logistics, along with cost estimation and environment setup, making this accessible for small teams seeking affordable computer vision solutions.

In this course, you learn about custom Chat Agents which are specialized AI assistants you configure with specific expertise, a defined persona, and access to your own reference documents and Spaces. Unlike My Assistant (general-purpose), a Custom Agent is laser-focused on a domain you define. By the end of this course, you will be able to implement chat agents so your team will have consistent, on-brand answers to common questions available 24/7 without relying on the availability of your most experienced colleagues.

This lab teaches you to integrate an AI agent with external tools using Amazon Bedrock AgentCore Gateway and the Model Context Protocol (MCP). You create a Gateway with JWT authorization, register Lambda functions as MCP tool targets, connect a pre-deployed nutrition coach agent, and validate end-to-end tool invocations using natural language queries.

This is a case-study course. Across eight steps it follows a retrieval-augmented question-answering system being stood up over a real 468-page corpus, and each step teaches one thing that carries into every RAG system you build afterwards: what a managed Knowledge Base actually owns, the two authorization layers that break many builds, what chunking does to your text, how to read raw retrieval before you let a model summarize it, what a citation does and does not prove, and how metadata filtering fixes a problem you can only see once you have measured it.

This course provides a technical deep dive into Amazon Elastic VMware Service (Amazon EVS). Designed for IT architects, VMware administrators, and cloud engineers, this course explores the architecture, networking, storage, security, migration, and operational considerations for deploying and managing VMware Cloud Foundation (VCF) environments on AWS. 

This assessment validates your comprehension of the Cloud Essentials Knowledge Badge Readiness Path. The assessment questions are based on the courses in this learning path. Take all of the courses and then verify your knowledge. Already have some knowledge on Cloud Essentials? Go directly to the assessment, test your knowledge. Score 80% or higher and earn an AWS Knowledge badge that you can share with your network and add to your resume. You can earn the Cloud Essentials Knowledge badge with a score of 80% or better. Badges are issued by Credly within 5-7 business days. Use this assessment as a method to continue building your knowledge. Assessment questions are randomized on each attempt which allows you to get more questions.

This assessment validates your comprehension of the Object Storage - Knowledge Badge Readiness Path. The assessment questions are based on the courses in this learning path. Take all of the courses and then verify your knowledge. Already have some knowledge on Object Storage? Go directly to the assessment, test your knowledge. You can earn the Compute Knowledge badge with a score of 80% or better. Badges are issued by Credly within 5-7 business days. Use this assessment as a method to continue building your knowledge. Assessment questions are randomized on each attempt which allows you to get more questions.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).* Learners will gain insight into how PAR Technology built a production-ready, multi-tenant natural language analytics agent on AWS that enforces strict data isolation. The course covers a three-layer security architecture -- cryptographic request signing with AWS SigV4, semantic validation on Amazon Bedrock, and programmatic data isolation via Split-Plane SQL -- designed to prevent cross-tenant data exposure even if the LLM is compromised. Participants will understand how to scope AI-generated SQL queries to the correct user permissions and datasets, enabling secure self-serve analytics at scale across diverse customer bases.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).* Learners can gain the ability to debug production AI agents that fail silently -- such as returning incorrect answers, entering infinite loops, or selecting wrong tools -- using Amazon Bedrock AgentCore Observability. The course covers how to use metrics, traces, and structured logs to follow agent reasoning steps, inspect tool invocations, and identify where execution diverges from expectations. Participants will learn structured workflows for resolving common failure patterns, moving beyond simple failure detection to understanding why failures occur, even when no explicit errors are raised.

This assessment validates your comprehension of the Amazon EKS Knowledge Badge Readiness Path. The assessment questions are based on the courses in this learning path. Take all of the courses and then verify your knowledge. Already have some knowledge on Amazon EKS? Go directly to the assessment, test your knowledge. You can earn the Amazon EKS Knowledge badge with a score of 80% or better. Badges are issued by Credly within 5-7 business days. Don't worry if you didn't achieve 80% or better. Use this assessment as a method to continue building your knowledge. Assessment questions are randomized on each attempt which allows you to get more questions.

Test Assessment (Do NOT Delete)

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).* Learners in this course will discover how to deploy ComfyUI workflows on Amazon SageMaker AI processing jobs to automate content generation at scale. They will learn to set up infrastructure using AWS CDK, configure GPU-accelerated processing, and generate hundreds of high-quality images in a single batch. The course covers practical, step-by-step guidance for scaling creative pipelines -- enabling enterprises to accelerate campaigns, produce on-brand visuals rapidly, and free creative teams from repetitive tasks so they can focus on high-impact strategy.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).* Learners will gain practical knowledge of building production-ready multi-tenant AI applications using Amazon Bedrock AgentCore. The course covers how to implement complete tenant isolation using native AWS capabilities, differentiate service tiers with minimal custom code, achieve granular cost attribution per tenant, and apply best practices for scalable multi-tenant AI architectures. Though demonstrated through healthcare AI agents serving multiple clinics, the patterns apply broadly to SaaS platforms, enterprise solutions, and managed services serving different customer organizations.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).* Learners will gain a practical understanding of how to secure AI coding agents -- such as Kiro and Claude Code -- within their development workflows. The course presents an application security control framework organized around two pillars: author-time controls (shaping what agents produce in the IDE) and build-time controls (gating what reaches production). Participants will learn to identify key risks like prompt injection and expanded tool access, and how to layer additional guardrails on top of existing SDLC practices so application security scales with agent-driven development across any toolchain or cloud environment.

In this course, you learn how to build a unified semantic layer that spans multiple normalized datasets using multi-dataset Topics in Amazon Quick Sight. You move from understanding the denormalization problem through designing, implementing, and consuming multi-dataset Topics that enable business users to ask natural language questions across your entire data model. After completing this course, you can design, build, and maintain a multi-dataset Topic in Amazon Quick Sight that enables business users to ask natural language questions spanning multiple normalized datasets — eliminating the need for denormalized 'big tables' and establishing centralized semantic governance.

This assessment validates your comprehension of the Storage Data Protection and Disaster Recovery Knowledge Badge Readiness Path. The assessment questions are based on the courses in this learning path. Take all of the courses and then verify your knowledge. Already have some knowledge on Data Protection? Go directly to the assessment, test your knowledge. The score report will identify your areas of strength and direct you to the courses where you can improve any knowledge gaps. You can earn the Storage Data Protection & Disaster Recovery Knowledge badge with a score of 80% or better. Badges are issued by Credly within 5-7 business days. Don't worry if you didn't achieve 80% or better. You can re-take the assessment after a 24-hour wait period. Use your score report to identify the courses where you can improve any knowledge gaps, then take the assessment again. Use this assessment as a method to continue building your knowledge. Assessment questions are randomized on each attempt which allows you to get more questions.

In this course, you will review the scope of the AWS Certified AI Business Strategist (AIB-C01 - English) exam, including the intended audience and exam topics. This course is part of 4 steps that you can use to prepare for your exam with confidence. To follow the 4 steps, enroll in the Exam Prep Plan: AWS Certified AI Business Strategist (AIB-C01 - English). The Exam Prep Plan includes exam-style questions, videos reviewing each exam domain and task statement, practice assessments, hands-on labs, flashcards, and more. The plan also includes role-based training to refresh your AWS knowledge and skills. Some of this content might require an AWS Skill Builder subscription.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).* Learners will discover how to build a custom incident triage assistant using Amazon Quick that orchestrates investigation and handoff steps in a single conversational workflow. By integrating the New Relic MCP Server and Asana, the agent can investigate incidents, assemble root cause analysis briefs with evidence links, and create tracked tasks -- all from a single prompt. Participants will learn how this approach reduces mean time to resolution, prevents knowledge loss between shifts, and establishes consistent investigation standards across on-call rotations.

This course concludes the Exam Prep Plan: AWS Certified AI Business Strategist (AIB-C01 - English) exam. This course is part of 4 steps that you can use to prepare for your exam with confidence. To follow the 4 steps, enroll in the Exam Prep Plan: AWS Certified AI Business Strategist (AIB-C01 - English) exam. Some of this content might require an AWS Skill Builder subscription.

This assessment validates your knowledge of Amazon WorkSpaces Core, including migration approaches and strategies, partner solution deployments such as Citrix and Omnissa Horizon on WorkSpaces Core, and additional partner offerings. Test your understanding of key concepts covered across the Amazon WorkSpaces Migrations Knowledge Badge Readiness learning path.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).* In this short course, AWS Partners learn how AWS-authorized distributors can accelerate their journey to building and scaling a successful AWS practice. You explore the role of distribution Partners, the key capabilities they offer — from funding and technical enablement to Marketplace support — and how to determine which engagement model aligns with your business goals. Whether you're just starting out or looking to grow an existing practice, this course helps you identify the right distributor resources to move faster, reduce operational complexity, and unlock new revenue opportunities.

Learn how the new AWS sign-up experience helps new users begin building faster with supported existing logins, AWS Builder ID, preconfigured projects, and simplified access to the AWS Management Console, AWS Settings, local development tools, and AI coding tools. This course also helps you recognize when your workload requires the advanced AWS experience for greater control over Regions, permissions, services, compliance, billing, or governance.

Learn how to develop AI strategies that align with business objectives by mapping AI capabilities to specific business outcomes and recognizing when a rule-based approach fits better. You will also weigh build, buy, and partner options against budget, timeline, and compliance requirements, decide which initiatives to scale, pause, or end, and evaluate what a transition to AI requires of data readiness, cost, and business continuity.

Learn how to establish the data and infrastructure foundations AI depends on by differentiating data readiness gaps in quality and accessibility and prioritizing the improvements that matter most, applying data strategy components such as ownership and sharing frameworks at enterprise scale, and identifying the technology gaps that would stop an initiative before it starts.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).* This 10-module product provides hands-on AWS service deep dives across advertising workflows—including ad insertion (MediaTailor), privacy-enhanced collaboration (Clean Rooms, Entity Resolution), real-time bidding infrastructure (RTB Fabric, MemoryDB/DynamoDB, NLB, EKS), and Generative AI for contextual targeting (Bedrock AgentCore, Elemental Inference) with IAB/GARM compliance. A final knowledge check validates learner comprehension across all nine content modules.

A closing reflection on what responsible AI engineering means for government, and the patterns, guardrails, and standards that make AI something your agency and citizens can trust.

This lab provides you with hands on experience on deploying and operating AI agents with Amazon Bedrock AgentCore CLI.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).* Learners will discover how AWS invented Parallel-EAGLE (P-EAGLE), a breakthrough method that parallelizes speculative decoding for large language models on Amazon SageMaker AI. The course explains how traditional approaches like EAGLE generate draft tokens sequentially -- creating latency that grows with speculation depth -- and how P-EAGLE eliminates this bottleneck by predicting all draft tokens simultaneously in a single forward pass using learnable placeholders. Participants will understand the architectural limitations of sequential drafting and how P-EAGLE transforms speculative decoding into a fully parallelized operation for improved inference throughput and reduced latency in production deployments.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).* Learners will gain insight into how Stripe built a production-grade AI agent system for financial compliance on AWS using Amazon Bedrock. The course covers the technical architecture of a ReAct agent framework, infrastructure decisions for a dedicated agent service, and the role of human oversight in maintaining accountability. Participants will also learn key lessons about task decomposition, orchestration patterns, prompt caching for cost optimization, and how to design agentic systems that scale compliance operations without compromising quality or auditability -- all drawn from a real-world deployment that reduced review handling time by 26 percent.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).* Learners will discover how to implement HippoRAG -- a retrieval augmented generation framework inspired by the human brain's hippocampal memory system -- using AWS services. The course covers building knowledge graphs with Amazon Neptune, applying Personalized PageRank for efficient graph traversal, using Amazon Bedrock for LLM capabilities, and Amazon Titan Embeddings for vector representations. Participants will learn how this approach overcomes standard RAG limitations by enabling multi-hop reasoning across multiple documents in a single retrieval step, suitable for enterprise-scale applications.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).* Learners will gain the ability to build an event-driven solution that automatically investigates AWS Systems Manager patch failures using AWS DevOps Agent. The course covers capturing patch failure events across multiple accounts and regions via Amazon EventBridge, enriching failures with SSM command output and instance metadata through AWS Lambda, and triggering autonomous root cause analysis that correlates logs, configuration state, and infrastructure changes. Participants will learn to deploy a centralized architecture that eliminates manual log correlation, reducing investigation time from hours to minutes for patch cycle failures spanning hundreds of managed nodes.

In this course, you will review Domain 1: AI Fundamentals and Literacy of the AWS Certified AI Business Strategist (AIB-C01 - English) exam. Prepare for the exam by exploring these topics and how they align to AWS services and to specific areas of study. Review content for each topic area of the domain, delivered by expert instructors. This course is part of 4 steps that you can use to prepare for your exam with confidence. To follow the 4 steps, enroll in the Exam Prep Plan: AWS Certified AI Business Strategist (AIB-C01 - English). Some of this content might require an AWS Skill Builder subscription.

This intermediate course equips experienced operations professionals with the skills to leverage Amazon Quick. Quick is the AI-powered digital workspace formerly known as Amazon QuickSight that transforms operational workflows, automate routine processes, and accelerate data-driven decision-making at enterprise scale. Learners will explore how the Amazon Quick platform integrates Quick Sight (business intelligence), Quick Flows (task automation), Quick Automate (complex process orchestration), Quick Research (deep research with cited reports), Quick Index (unified enterprise knowledge), and custom Chat Agents into a cohesive operational toolkit. Through real-world operations scenarios, you will build operational dashboards with SPICE-accelerated datasets, design natural language automation workflows for procurement and supply chain processes, conduct deep operational research across enterprise and third-party data sources, and implement governance controls for secure, scalable deployments.

In this course, you will review Domain 2: AI Strategy and Business Value Creation of the AWS Certified AI Business Strategist (AIB-C01 - English) exam. Prepare for the exam by exploring these topics and how they align to AWS services and to specific areas of study. Review content for each topic area of the domain, delivered by expert instructors. This course is part of 4 steps that you can use to prepare for your exam with confidence. To follow the 4 steps, enroll in the Exam Prep Plan: AWS Certified AI Business Strategist (AIB-C01 - English). Some of this content might require an AWS Skill Builder subscription.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/aws-training).* Agentic AI coding tools like **Kiro** move past autocomplete. Kiro is an agentic IDE and CLI, built on AWS, that turns prompts into specs, code, docs, and tests. To do that work it reads and writes files, runs commands, calls external tools through the **Model Context Protocol (MCP)**, and reaches into AWS resources. That reach is exactly what makes it productive for developers, and exactly why a careless enterprise rollout is dangerous. This course is the governance view: what an administrator, a security reviewer, and a developer working inside these controls all need to know before Kiro goes org-wide. The course opens by framing why agentic coding changes the governance conversation. Spec-driven development with Kiro redistributes effort toward upfront planning and structured review, but the same speed that makes a team productive also raises the cost of a wrong action if boundaries are not in place first. Many organizations start with agentic coding as an individual experiment and then move to production at enterprise scale, and that transition is where governance becomes necessary rather than optional. The core of the course maps every governance concern to the control that addresses it, using a running question to test each one: **can a developer simply turn this control off**. **Identity and access integration** authenticates enterprise users through AWS IAM Identity Center, Okta, or Microsoft Entra ID, with subscriptions assigned to users and groups rather than ad hoc logins, scoped by a Kiro profile of AWS account plus Region. **Subscription and license management** tracks credit and overage consumption per user from the same console. **MCP server governance** lets administrators disable MCP entirely or publish an HTTPS-hosted allow-list of approved servers that clients sync at startup and every 24 hours, terminating any locally installed server that drifts off the list, and failing closed if the client cannot reach the governance API. **Model governance** lets administrators curate an approved model list and set a default, which is curation rather than bring-your-own-model, with direct implications for data residency since generally available models use regional inference while experimental models may route globally. **Feature controls** add broader switches, including a web tools toggle that disables web search and fetch organization-wide. The course then covers how governance creates visibility. The native usage dashboard shows aggregate metrics. A per-user activity report writes a daily CSV to an S3 bucket in your own account, and prompt logging records actual prompts and responses to your own bucket for audit and compliance, which means the data lives in your account and can feed your own Athena and QuickSight dashboards. On data handling, enterprise use means Kiro does not use your content for service improvement or model training and opts you out of telemetry automatically, and encrypts in transit and at rest by default. Content itself may still be stored in the Region where your profile is configured in order to provide the service, which is how prompt logging and activity reports operate. That leaves two real decisions: a customer-managed KMS key, and residency for experimental models. The course closes on what lives outside the central console. **Steering files** give Kiro persistent rules as soft guidance. **Hooks** are enforceable: a PreToolUse command hook that exits with any non-zero code blocks the tool invocation outright, regardless of how the prompt was phrased. **Checkpoints and rewind** provide session recovery, with the important limit that checkpoints do not track changes made by MCP tools or by bash commands the agent runs. The course ends on the boundary that holds regardless of any client-side setting: Kiro holds no cloud credentials of its own and acts with whatever AWS credentials already exist in the developer's environment, so least-privilege IAM and short-lived sessions cap the blast radius no matter what a prompt, hook, or model attempts, because AWS enforces that boundary server-side. This distinction is load-bearing, because MCP and model governance are enforced client-side and can be circumvented by a user with local administrative access.

Learn how to lead enterprise-wide AI change by building executive sponsorship and champion networks, and assembling cross-functional teams with real accountability rather than standing meetings. You will also communicate openly about AI's effect on jobs, diagnose which cultural barrier is blocking adoption before choosing an intervention, match workforce development to different employee segments, and lead role transitions that balance AI capabilities against human strengths.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).* Learners will discover how A2A -- a major Italian energy and utilities company -- built a FinOps platform on AWS to normalize cost and usage data across AWS, Azure, and Google Cloud. The course covers architectural decisions, dashboard implementations, and the transition from showback to chargeback models for financial accountability. Participants will gain transferable lessons for consolidating multi-cloud billing data, reducing manual budget processes, identifying cost savings, and providing actionable spending visibility by business unit -- all applicable to organizations managing complex, multi-cloud environments at scale.

This course provides a comprehensive guide to maximizing Amazon EKS cost-efficiency and operational simplicity through three complementary technologies: EKS Auto Mode, AWS Graviton processors, and Amazon EC2 Spot Instances. You will learn to automate cluster infrastructure, migrate workloads to Graviton, handle Spot interruptions gracefully, and combine all three for maximum savings.

Learn how to use Amazon Bedrock AgentCore Evaluations to assess, measure, and improve AI agent performance. In this course, you’ll explore evaluation concepts, configure evaluators, run offline and online evaluations, interpret results, and apply evaluation insights to build more reliable agentic applications.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).* Learners will gain the ability to send on-premises firewall, router, and switch syslog directly into Amazon CloudWatch using managed syslog ingestion -- eliminating the need to build and maintain a separate EC2-based collection tier. They will learn how CloudWatch parses common syslog formats, how to query extracted fields, and how to create log-based alarms on critical events. Finally, they will discover how to route those alarms to the AWS DevOps Agent so it can automatically investigate issues using the device logs, providing end-to-end on-premises visibility without extra infrastructure.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).* Learners in this course will gain the ability to diagnose AI agent failures quickly using the Strands Evals SDK's detector functions. They will learn to interpret structured output -- including categorized failures with confidence scores, causal chains linking root causes to symptoms, and fix recommendations indicating whether changes belong in system prompts or tool definitions. The course also covers integrating detection into evaluation pipelines for automated diagnosis on every test run, reducing diagnosis time from hours to minutes and moving beyond simple scores to understand why agents fail and how to fix them.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).* Learners will discover how Pelago built and deployed a serverless AI assistant in two weeks using AWS services like Amazon Bedrock and AWS Lambda. The course covers designing an event-driven architecture that generates contextually aware suggestions for healthcare coaches -- preserving human-in-the-loop oversight while scaling personalized patient interactions. Participants will gain insights into overcoming constraints like long-term conversation context, rapid development timelines, and infrastructure simplicity, all within a real-world digital health setting focused on substance use disorder support.

This assessment validates your comprehension of the Serverless Knowledge Badge Readiness Path. The assessment questions are based on the courses in this learning path. Take all of the courses and then verify your knowledge. Already have some knowledge on Serverless? Go directly to the assessment, test your knowledge. You can earn the Serverless Knowledge badge with a score of 80% or better. Badges are issued by Credly within 5-7 business days. Use this assessment as a method to continue building your knowledge. Assessment questions are randomized on each attempt which allows you to get more questions.

Introduction to Amazon Quick Desktop for retail manager. You learn how to connect your local folders, knowledge graph insights, and additional tools to manage context for your operations. You can set Amazon Quick Desktop to take actions while you sleep.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).* Learners will discover how bitdrift scaled to 121 million concurrent gRPC connections during live T20 World Cup cricket broadcasts using Amazon CloudFront. The course covers how DNS routing policy choices -- specifically weighted versus multi-value answer routing -- can cause connection concentration on a single origin endpoint under extreme traffic surges. Participants will learn how bitdrift and the AWS account team diagnosed this imbalance and resolved it with a single DNS configuration change, enabling their multi-NLB architecture to distribute load evenly and serve all devices with zero server-side errors.

This is a build-along course. You will write one program in six passes, and each pass teaches one thing that carries into every Bedrock application you write afterwards: the request/response shape, streaming, conversation state, token accounting, prompt caching, and finally consolidating all of it into a tool you would actually use.

This assessment validates your knowledge of the topics covered in the End User Computing on AWS Knowledge Badge Readiness Path. After passing this assessment, the trainee will be granted the End User Computing on AWS accreditation. This training consists of an online assessment composed of 60 questions. You require a score of 80 percent or greater to pass. You have unlimited attempts to pass the assessment.

Learn how to use the built-in visual file editor in AWS CloudShell to make focused changes to scripts, configuration files, infrastructure templates, and other text-based files. You’ll practice the edit, save, verify, run, and review workflow while learning how the visual editor works alongside existing CloudShell tools, permissions, storage, and execution behaviors.

Earners of this badge have demonstrated the ability to configure and operate Amazon Connect's ML-powered workforce management capabilities, including generating contact volume and handle time forecasts, translating forecasts into staffing capacity plans using service level optimization, creating and publishing optimized agent schedules, monitoring real-time schedule adherence, managing intraday operations, and driving continuous improvement through FCS metrics and analytics.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).* This 7 module product covers covers the AWS Unified Operations Support Model (team structure, support catalogue entitlements, and engagement guidance for advertising workloads) alongside advertising-specific AWS resources—including the Video Streaming Advertising (VSA) Lens and Streaming Media Lens for Well-Architected assessments, public workshops and GitHub solutions for hands-on exploration, and Solution Guidance for production implementation of advertising workloads.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).* Learners will discover how to move coding agents off local laptops and onto Amazon Bedrock AgentCore, a cloud runtime that provides each agent session with an isolated Linux microVM, persistent workspace, and real shell. The course covers the five requirements any coding agent needs -- a shell, filesystem, checked-out project, dependencies, and permissions -- and shows how AgentCore delivers identity management, a unified MCP gateway for tools like GitHub and Slack, and built-in observability via CloudWatch. Participants also learn to benchmark multiple agents side by side on latency and cost.

In this intermediate-level course, you learn how to build real-time data streaming solutions using Amazon Kinesis Data Streams. You explore the core concepts of Kinesis Data Streams, including how to create and configure data streams, produce and consume records, and design architectures that process streaming data at scale. Through practical guidance, you gain the skills needed to implement Kinesis Data Streams in your AWS environments, enabling you to handle continuous data ingestion and real-time analytics for your applications.

Learn how to put generative AI to work by applying prompt engineering for reliable output, recognizing when token limits and context windows affect performance, and knowing when RAG or fine-tuning is the right adaptation technique for a business need.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).* Learners will gain practical guidance on training multi-turn reinforcement learning agents in Amazon SageMaker AI. The course covers how to build trustworthy training environments, set up external evaluation, design task-aligned rewards, manage complexities across multiple turns, and monitor key metrics for iteration. Drawing on the SOP-Bench benchmark across 12 business domains, it addresses challenges like reward hacking and corrupted training signals. Learners will also understand SageMaker AI's modular agent-environment interface, serverless execution, and integration options -- equipping them to reliably train agents that handle dependent steps such as tool calls, error recovery, and decision-making.

AWS Glue 6.0 is the newest version of the AWS Glue Spark runtime, and it is a major release. It moves the engine to Apache Spark 4.2.0, brings Iceberg format version 3 to AWS Glue, adds two new ways to build pipelines (Spark Declarative Pipelines and real-time mode for streaming), and removes several long-deprecated pieces of the runtime along the way. This course covers what is new in AWS Glue 6.0 , why it matters, and what you have to do to your existing jobs before they will run on it. This is a concepts course: every code example is shown inline, and you do not need an AWS account to follow along. Prerequisite: No prior AWS Glue 6.0 experience is assumed. Familiarity with the basic idea of an AWS Glue Spark ETL job (a script, an AWS Glue version, worker types, and job parameters).

The Official Practice Question Set: AWS Certified AI Business Strategist (AIB-C01 - English) consists of 20 questions. This question set aligns with the (AIB-C01 - English) version of the exam and exam guide. Official Practice Question Sets feature 20 questions developed by AWS to demonstrate the style of Certification exam questions. Each question includes detailed feedback for answer choices, including recommended resources to deepen your understanding of key topics. You can take the question set multiple times. Each time will include the same questions in a different order. This course is part of 4 steps that you can use to prepare for your exam with confidence. To follow the 4 steps, enroll in the Exam Prep Plan: AWS Certified AI Business Strategist (AIB-C01 - English). Some of this content might require an AWS Skill Builder subscription.

In this lab, you utilize Amazon Elastic Container Service (Amazon ECS) to troubleshoot and resolve task misconfiguration failures in a containerized application environment.

In this lab, you deploy a containerized application by using Amazon Elastic Kubernetes Service (Amazon EKS).

This course is designed to help you learn how to use Amazon Quick for Healthcare. Quick consolidates fragmented healthcare data into a unified AI-powered intelligence platform that enables clinical leaders to identify performance gaps, surface patient safety risks, and act on real-time insights without technical expertise.

This course introduces you to Amazon Quick Apps, a no-code application builder that lets you create custom internal tools, customer-facing portals, and operational applications directly within Amazon Quick. You learn how to design, build, share, and iterate on apps without writing code, submitting IT tickets, or waiting for engineering resources. Amazon Quick Apps change how internal tools get built. Instead of waiting months for IT to build a custom tracker, portal, or dashboard, you describe what you need and Quick builds it. You learn to use this capability for immediate operational impact with zero engineering resources required.

Use Amazon Bedrock AgentCore Observability with CloudWatch Metrics and Log Analytics to monitor, diagnose, and fix performance issues in a production AI agent deployed on AgentCore Runtime.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/aws-training).* Most engineering teams already have an AI assistant. Fewer have worked out how to make it behave like a colleague who knows their codebase, their conventions, and their cloud account. That gap is what this course is about. **Kiro** is an agentic AI development environment available across three surfaces: the **Kiro IDE**, **Kiro Web**, and the **Kiro CLI**. This course walks through the capabilities that let you shape how a Kiro agent works, then shows the same agent applied to three different jobs by three different personas. The first half covers the mechanics. You will see how Kiro handles context: native **Model Context Protocol (MCP)** integration to reach docs, databases, and APIs; **steering files** to set how agents behave per project; and image input, so a photo of a whiteboard architecture session or a screenshot of a UI design can guide implementation. You will see **agent hooks**, which delegate work to agents that fire on events such as a file save and run autonomously in the background to generate documentation, write unit tests, or optimize performance. On the CLI side, **custom agents** narrow an agent to a specific job with pre-approved tool permissions, context files, and prompts, while **headless execution** runs Kiro programmatically inside a CI/CD pipeline with no user input, authenticating via API keys, device codes, or PKCE. The course then covers **Kiro powers**, launched at re:Invent 2025, which extend agent capabilities through sharable best practices. A power can bundle MCP servers for access, steering files for context, and hooks for actions. Powers come from ISVs such as Figma, Postman, and Stripe, from domain experts packaging specialized use cases, and from community members sharing their own from GitHub. Alongside powers you will see how **skills** work: a folder containing a SKILL.md with YAML frontmatter and Markdown instructions, loaded progressively so only lightweight metadata sits in context until a task matches the skill's description. The section closes with a side-by-side comparison of **skills**, **steering**, **and MCP**: what each provides, when each loads, what triggers it, and what each is best for. They are complementary rather than competing. The second half is applied, one section per persona. **Developer.** Two scenarios: asking Kiro to locate where item validation happens during checkout in an unfamiliar project, made faster by steering; and asking Kiro to deploy an application into an AWS account. **DevOps.** Reviewing an Amazon ECS service and an Amazon RDS instance for right-sizing, using subagents running in parallel. **Security Operations.** The largest section, built on a deliberately small architecture: a single SKILL.md acting as a security operations analyst persona that routes to three workflows, sharing one definition of the MCP data source, grounding discipline, safety rules, and a standard ticket-ready output format. The pattern is three steps: MCP gathers the evidence by calling live AWS APIs, the skill reasons over it using expert workflow logic, and Kiro drafts the work product you would otherwise write by hand. It is read-only by default and confirms before any mutating action. Three use cases are presented and then run live:• **Investigate security events** with Amazon GuardDuty: pull the finding, enrich the actor identity, scope blast radius through AWS CloudTrail, classify true or false positive, recommend containment, and draft a triage note• **Remediate weak configuration** with AWS Security Hub CSPM: pull failed controls, prioritize by severity and exposure, confirm resource state, propose the fix, apply it, then re-check that the control actually passed• **Implement least privilege** with AWS IAM Access Analyzer: pull unused-access findings, review real CloudTrail activity, generate a refined policy from what the role actually did, produce a current-versus-proposed diff, and draft guidance for the resource owner One caveat is stated plainly in the session: this is an analyst assistant. It does not replace your SIEM and it does not perform real-time streaming detection. Its value is in the investigation, remediation, and documentation work that surrounds a finding that already exists. The course closes with the **Agentic AI Security Scoping Matrix**, a four-scope model spanning human-initiated with no agency through automated initiation with full agency, used to reason about how security requirements scale as you hand an agent more autonomy, and where current agentic AWS tooling, Kiro included, sits on it.

Learn how to measure and demonstrate AI business value by defining KPIs that capture both tangible gains and intangible ones like customer satisfaction, and establishing baseline metrics before launch so improvement can be proven rather than assumed. You will also calculate ROI across time savings, cost reduction, and productivity gains, identify leading indicators that predict success, and control the cost drivers that grow after deployment.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).* Learners will discover how to use the Amazon Bedrock Guardrails InvokeGuardrailChecks API to apply targeted safety controls at any point in agentic AI applications without creating separate guardrail resources. They will learn how the API operates in detect-only mode, returns numeric scores for each safeguard, and enables custom thresholds and actions -- such as blocking, bypassing, retrying, or logging results. The course covers building safe multi-turn agentic workflows where each step carries different risk profiles, giving developers flexible, lightweight safety integration throughout the agentic loop.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).* Learners will gain a practical understanding of how to match the right AI tool to specific FinOps use cases on AWS. The course covers five tools -- AWS FinOps Agent, Amazon QuickSight, Kiro, Amazon Q (In Console), and AWS DevOps Agent -- explaining each tool's core purpose and ideal scenarios. Participants will learn to distinguish between purpose-built managed solutions and general-purpose tools, understand trade-offs in accuracy and operational readiness, and make informed decisions that accelerate their FinOps practice without adding unnecessary complexity.

The Official Practice Question Set: AWS Certified Machine Learning - Associate (MLA-C02 - English) consists of 20 questions. This question set aligns with the (MLA-C02 - English) version of the exam and exam guide. Official Practice Question Sets feature 20 questions developed by AWS to demonstrate the style of Certification exam questions. Each question includes detailed feedback for answer choices, including recommended resources to deepen your understanding of key topics. You can take the question set multiple times. Each time will include the same questions in a different order. This course is part of 4 steps that you can use to prepare for your exam with confidence. To follow the 4 steps, enroll in the Exam Prep Plan: AWS Certified Machine Learning - Associate (MLA-C02 - English). Some of this content might require an AWS Skill Builder subscription.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).* Learners will discover how Kiro CLI -- an AI-powered command-line assistant with Model Context Protocol integration -- streamlines AWS support case workflows. The course covers combining CloudWatch investigation, AWS documentation lookup, and support case creation into a single conversational interface. Through three real-world scenarios -- AWS Glue job failures, Lambda cold starts, and WAF false positives -- learners see how to eliminate manual log collection, context switching, and repetitive documentation tasks, enabling operations teams to resolve issues faster while maintaining thorough records.

PostgreSQL for Amazon Aurora and RDS Advanced Concepts Assessment

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).* Learners will gain hands-on experience building an automated healthcare claims processing pipeline using Amazon Bedrock and AWS HealthLake. The course covers using Amazon Bedrock Data Automation for intelligent document extraction from CMS-1500 claim forms, deploying an AI agent with Strands Agents on Amazon Bedrock AgentCore to validate extracted data against patient and provider records, and creating standardized FHIR resources. Participants will learn to connect these services into an end-to-end workflow -- triggered by S3 uploads and orchestrated through Lambda -- that reduces manual processing while maintaining accuracy through automated validation and SNS notifications.

Build an orchestrated, incremental, multi-source data pipeline with AWS Glue. Crawl two raw datasets in Amazon S3, join them in a Glue Visual ETL job, orchestrate the crawler and job as a single Glue Workflow driven by triggers, process new data incrementally with job bookmarks, and visualize the result in a Glue interactive notebook. Validate with Amazon Athena.

In this intermediate-level course, you explore how to build solutions with Amazon OpenSearch Service. You learn how to set up and configure OpenSearch domains, ingest and index data, and perform powerful search and analytics queries. Through practical guidance, you gain the skills to design and implement OpenSearch-based architectures that support log analytics, full-text search, and real-time application monitoring. By the end of this course, you will have the foundational knowledge to confidently build and manage OpenSearch Service workloads in your AWS environment.

In this course, you will review Domain 2: ML Model and Foundation Model (FM) Development of the AWS Certified Machine Learning Engineer - Associate (MLA-C02 - English) exam. Prepare for the exam by exploring these topics and how they align to AWS services and to specific areas of study. Review content for each topic area of the domain, delivered by expert instructors. This course is part of 4 steps that you can use to prepare for your exam with confidence. To follow the 4 steps, enroll in the Exam Prep Plan: AWS Certified Machine Learning Engineer - Associate (MLA-C02 - English). Some of this content might require an AWS Skill Builder subscription.

A short welcome to the learning plan for the engineers, developers, architects, and data specialists building the next generation of public sector services, with a look at how you'll learn to build and operate generative AI on AWS securely, reliably, and at scale.

This assessment validates your knowledge of the topics covered in the AWS Transform Fundamentals Knowledge Badge Readiness Path learning plan. After passing this assessment, the trainee will be granted the AWS Transform Fundamentals accreditation. This training consists of an online assessment composed of 60 questions. You require a score of 80 percent or greater to pass. You have unlimited attempts to pass the assessment.

This assessment validates your comprehension of the Storage Data Migration Knowledge Badge Readiness Path. The assessment questions are based on the courses in this learning path. Take all of the courses and then verify your knowledge. Already have some knowledge on Storage Data Migration? Go directly to the assessment, test your knowledge. You can earn the Storage Data Migration Knowledge badge with a score of 80% or better. Badges are issued by Credly within 5-7 business days. Don't worry if you didn't achieve 80% or better. Use this assessment as a method to continue building your knowledge. Assessment questions are randomized on each attempt which allows you to get more questions.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).* Learners will discover how Amazon Corporate Infrastructure Services built a Full Stack Observability platform using Amazon OpenSearch Serverless to monitor critical services across 400 offices in 50+ countries. The course covers how the team overcame fragmented visibility, data silos, and slow issue detection by integrating multiple monitoring sources into a unified platform. Participants will learn the technical architecture, implementation approach, and lessons learned -- including how the solution achieved a 5-minute mean time to detect, saved thousands of engineering hours annually, and maintained 99.9% platform uptime.

*This course was developed by members of AWS Technical Field Communities (TFC), an AWS community of technical experts. The content is intended to complement our standard training curriculum and augment your AWS learning journey. We are aware some courses have accessibility limitations and are working to address. If you require accommodation, please contact [AWS Training and Certification Customer Support](https://support.aws.amazon.com/#/contacts/one-support?formId=trainingCertification).* Learners will understand how AWS Shield Advanced is transitioning to the AWS WAF Anti-DDoS managed rule group as its default -- and eventually only -- application-layer DDoS protection. The course covers what the new rule group offers, including faster traffic baselining, rapid attack response, and Challenge actions driven by suspicion-level labels. Participants will learn the migration timeline, how the rule group is added in Count mode without disrupting traffic, what steps to take before the finish date, and how monitoring and metrics will change throughout the transition phases.