Skip to content

Customer story · Hospitality technology

Mr Yum: building and operating Kubernetes on AWS.

CloudHero built and maintained a Kubernetes cluster on AWS for Mr Yum’s digital menu and table-ordering platform. The engagement brought senior DevOps architecture and engineering into both platform delivery and ongoing operations.

Mr Yum logo
A historical CloudHero engagement with Mr Yum.
Product
Digital menus and table ordering
Platform
Kubernetes and AWS
Engagement
Build and ongoing maintenance

The platform

Infrastructure behind digital menus and table ordering.

Mr Yum’s product brought menus and ordering into a web-based experience for hospitality venues. CloudHero’s work focused on the infrastructure supporting that product: building a Kubernetes cluster on AWS and continuing to maintain it.

That scope joined platform delivery with operations. The engagement combined senior DevOps architecture and engineering, giving the work both architectural direction and hands-on implementation capacity.

For a digital ordering service, the useful unit of reliability is the customer journey. A running cluster is one part of that journey; databases, event streams, application changes and operational visibility also need attention.

Engineering scope

A Kubernetes platform with data and delivery dependencies.

The engagement’s technology stack included Amazon EKS, Kubernetes and Istio, together with Terraform, GitLab and Buildkite. It also included AWS data and infrastructure services, CockroachDB and Datadog. Grouping them by responsibility makes the platform easier to understand than a single list of tool names.

Application platformAmazon EKS · Kubernetes · Istio
Infrastructure and deliveryTerraform · GitLab · Buildkite
Data and event servicesRDS · CockroachDB · Kinesis · MSK
Compute, storage and visibilityAmazon EC2 · S3 · Datadog

Technology groups from CloudHero’s original engagement account. This is an overview of the recorded stack, not a network topology or a claim that every component ran inside Kubernetes.

What CloudHero delivered

CloudHero built and maintained the AWS-hosted Kubernetes cluster. The work continued beyond establishing the initial environment, bringing infrastructure engineering into the ongoing operation of the platform.

The data services and delivery tools illustrate the breadth of the environment around that cluster. Platform decisions need to account for how applications are released, how they connect to data, and how engineers investigate problems across those boundaries.

From build to operation

Make the operating responsibilities part of the platform design.

The central lesson from this engagement is continuity: the team building the environment also contributed to maintaining it. For teams planning a comparable platform, the handover and support model deserve attention while the architecture is still being designed.

When changing the platform

Agree how infrastructure changes are reviewed, how application compatibility is checked, and how a release can be rolled back. Include the delivery pipeline and dependent services in that review.

When running the platform

Define who owns cluster maintenance, access, observability and incident coordination. Make the boundary between platform operations and application ownership clear to both teams.

AWS documents a shared responsibility model for Amazon EKS. A managed service still leaves workload and configuration responsibilities with the customer. The scope of the engagement should make those responsibilities concrete.

Questions for your team

What should you agree before commissioning an EKS platform?

  1. Choose the acceptance journey. Which user actions must work before a release or platform change is considered successful?
  2. Map data dependencies. Which databases, object stores and event services does each application need, and who owns them?
  3. Set the release process. Who approves a deployment, watches the result and decides when to roll it back?
  4. Plan maintenance. How will cluster and workload changes be tested and scheduled alongside product delivery?
  5. Agree operational ownership. Who receives an alert, investigates it and coordinates application and infrastructure teams?

CloudHero can use these questions to scope a Kubernetes and Amazon EKS engagement, or provide ongoing platform operations for an existing environment.

Case-study basis: CloudHero’s original Mr Yum engagement account and project records. This historical story describes delivered scope and engineering responsibilities; it does not present an uptime, latency or cost benchmark.

Your next step

Plan the cluster and the work of running it.

Review your AWS architecture, application dependencies and support needs with engineers who build and operate Kubernetes platforms.

Contact CloudHero

Tell us what you're working on

Our sales team replies by email within 1 business day.

Topic

We use your details only to reply. Privacy policy