What is AWS re:Invent?
re:Invent is the big AWS conference every December in Las Vegas. It is where AWS dumps a year’s worth of launches, keynotes, and roadmap signals all at once. Hundreds of announcements, a lot of noise, and usually a handful of things that actually matter to people running production. This is one of the ones that matters.
The announcement: Amazon EKS Capabilities. It is an extensible set of Kubernetes-native platform features for workload orchestration, Kubernetes resource composition, and managing AWS cloud resources from inside the cluster. The interesting part is not the feature list. It is that AWS runs these fully managed, and they are built on open source tools many of us already run by hand: Argo CD, AWS Controllers for Kubernetes (ACK), and Kube Resource Orchestrator (KRO).
Here is the operational shift. Instead of installing these controllers inside your clusters and owning their lifecycle, AWS operates them in EKS service-owned accounts. That means scaling, patching, and upgrading those foundational components stops being your problem.
What AWS Announced
EKS Capabilities are managed, Kubernetes-native features covering continuous deployment, resource orchestration, and cloud resource management. Each one is optional and enabled independently per cluster. You turn on what you need, you skip what you do not.
Three capabilities at launch.
Argo CD
The declarative GitOps tool for continuous deployment on Kubernetes. If you have done GitOps on Kubernetes, you have probably already run it. It is not niche: more than 45 percent of end users surveyed by the Cloud Native Computing Foundation in 2024 reported using it.
AWS Controllers for Kubernetes (ACK)
ACK gives you custom Kubernetes resources for managing AWS cloud resources straight from the cluster. Provision an S3 bucket or an RDS instance with the same kubectl apply you use for everything else. Plenty of enterprise platform teams already run ACK to keep cloud and cluster workflows in one place.
Kube Resource Orchestrator (KRO)
KRO lets you define reusable resource bundles and abstractions, so the way custom resources get created and consumed stays simple without drifting away from native Kubernetes patterns. Think of it as the glue for packaging up the messy multi-resource setups your developers should not have to assemble by hand.
Why EKS Capabilities Matter
1. Less Operational Overhead for Platform Teams
These capabilities run in AWS managed accounts, not inside your clusters. Nobody on your team installs them, scales them, patches them, or babysits their upgrades. AWS owns the underlying components and the reliability, security, and lifecycle work that comes with them. That is real time back, especially the kind of work that tends to land on whoever is on call.
2. Streamlined Kubernetes Application Delivery
Your developers keep working the way they already do. They apply declarative config with kubectl or through pipelines, and these capabilities extend that with native integrations for deployment and resource composition. No new mental model to learn, no parallel toolchain to maintain.
3. Integrated Support for Building and Scaling Platforms
Each capability works on its own or together. Combined, they cover:
- GitOps-driven continuous deployment
- Cloud resource provisioning from Kubernetes
- Composition and orchestration of complex resource bundles
That is most of the internal platform people end up building from scratch, except you are not building it from scratch.
How EKS Capabilities Work
Platform engineers and cluster admins decide which capabilities to enable per cluster. EKS runs the infrastructure and exposes cluster-scoped resources that developers hit through standard Kubernetes APIs.
For developers, nothing changes day to day. They create and modify Kubernetes resources declaratively, and the enabled capabilities handle the higher-level automation and orchestration behind the scenes.
Additional Considerations
Permissions
EKS Capabilities are configured through AWS Identity and Access Management. Some include optional single sign-on, such as Argo CD wired up to IAM Identity Center.
Upgrades
EKS updates enabled capabilities and their dependencies automatically. It analyzes for breaking changes, patches components, and surfaces problems through EKS cluster insights, so you are not chasing a controller upgrade that quietly broke something.
Resource Adoption
ACK can adopt existing AWS resources into ACK management. It also supports read-only resources, which gives you a path to migrate gradually from Terraform or AWS CloudFormation instead of forcing a hard cutover.
Availability and Pricing
EKS Capabilities are available in commercial AWS Regions. No upfront commitments, no minimum fees. You pay for the capabilities and resources you use.
A More Integrated Kubernetes Experience
With EKS Capabilities, AWS is pushing the managed Kubernetes experience past control plane operations. By pulling popular open source tools and cloud resource controllers into a fully managed environment, it cuts the operational complexity platform teams usually absorb and gives them a more cohesive base for building and scaling on Kubernetes.
The result is a single model for continuous deployment, resource orchestration, and cloud integration. Less time keeping the plumbing alive, more time shipping the applications that actually justify the cluster.
If you need support securing, governing, or scaling your Kubernetes environments on AWS, our managed cloud specialists are here to help. Contact us to get started.