AWS

About AWS

AWS is the cloud-services portfolio from Amazon Web Services. It spans infrastructure, databases, networking, analytics, security, developer tooling, artificial intelligence, machine learning and business applications. This page focuses on how the AWS range is organized and how buyers should think about combining services. Company-level financial information belongs on the Amazon Web Services company page, while service-specific pricing and capabilities belong on individual product pages.

How is the AWS portfolio organized?

AWS is not sold as one monolithic application. Its catalogue is organized into service families that customers combine to build an architecture. Core categories include compute, storage, databases, networking and content delivery, analytics, security and identity, developer tools, management and governance, migration, artificial intelligence and machine learning, and business applications. AWS states that more than 200 services are available across its cloud platform.

Well-known services include Amazon EC2 for compute, Amazon S3 for object storage, Amazon RDS and Aurora for relational databases, Amazon DynamoDB for NoSQL, AWS Lambda for serverless computing, Amazon EKS for Kubernetes, Amazon CloudFront for content delivery, Amazon Redshift for analytics, Amazon SageMaker AI for machine learning and Amazon Bedrock for generative AI. These services are designed to be composed rather than treated as isolated products.

What connects AWS services across the platform?

AWS services share common infrastructure and account-level controls. Identity and Access Management, Organizations, CloudTrail, CloudWatch, networking, encryption and tagging can span many workloads and services. This shared control plane is useful for organizations that want one governance model across infrastructure, data and application services.

The same interconnected model creates operational responsibility. A team that adopts many AWS services must manage permissions, accounts, network boundaries, logs, service quotas, Regions, encryption keys and cost allocation consistently. Architecture quality therefore depends on how the services are combined, not only on the capabilities of each product.

How does AWS pricing work across the portfolio?

There is no single AWS subscription price. Most services use consumption-based billing, and the unit varies by service. Compute can be billed by instance time or capacity, object storage by stored data and requests, databases by capacity and storage, network services by traffic, and AI services by tokens, requests, model capacity or specialized tools.

AWS also offers pricing mechanisms such as Savings Plans, reserved capacity, free-tier allowances and negotiated enterprise agreements. Buyers should model the full workload rather than compare only one service's headline rate. Data transfer, backups, observability, idle resources, high availability and support can materially change total cost.

How does AWS handle global deployment?

AWS currently operates 39 geographic Regions and more than 120 Availability Zones, with additional Regions and Availability Zones announced. A Region is a separate geographic area, while Availability Zones are isolated infrastructure locations within a Region. This lets customers choose deployment locations for latency, resilience, regulatory and data-residency needs.

Service availability can differ by Region, so an architecture that works in one location may not be identical in another. Organizations planning multi-region systems should confirm service support, quotas, replication behavior and pricing before standardizing a design.

Where does AI fit in the AWS range?

AI now spans several parts of the AWS portfolio. Amazon Bedrock provides managed access to foundation models and generative-AI tooling. Amazon Bedrock AgentCore is focused on building and operating AI agents. Amazon Nova is AWS's foundation-model family, while Amazon SageMaker AI covers broader model development, training and deployment workflows.

AWS also embeds AI into business and developer services. Buyers should separate model access, agent infrastructure, machine-learning development and end-user AI applications because they solve different problems and can use different pricing models.

Who is the AWS portfolio best suited for?

AWS is a strong fit for organizations that need broad infrastructure choice, global deployment, managed databases, data platforms, developer services or AI under one cloud provider. It can support small applications as well as complex enterprise environments, and its large service catalogue gives architecture teams many ways to combine managed and lower-level building blocks.

The trade-off is complexity. Teams with limited cloud expertise may find a narrower managed platform easier to operate. Organizations already standardized on Microsoft Azure, Google Cloud or another platform should compare migration cost, existing skills, contractual commitments, identity systems and data-transfer requirements before adding another major cloud.

What should buyers check before standardizing on AWS?

Before broad adoption, buyers should map workloads to Regions, define account and identity structure, set cost-allocation rules, decide which services are approved, establish logging and security baselines, and understand support requirements. They should also identify where architecture choices create service-specific lock-in, especially around proprietary databases, event systems, serverless patterns and AI tooling.

A useful AWS decision is therefore not simply whether the platform has a required feature. It is whether the organization can govern, operate and financially manage the combination of services needed for that workload.