This is very basic API template project; which gets hosted in Kubernetes Cluster. API Project is structured as below

a) SeedWebApi: Host shell; which exposes api endpoints
b) ServiceLayer: All the business logic; processed using Domain objects 
c) PersistenceLayer: Uses In-Memory database to store data

a) Installed Azure Powershell CLI (This is required to connect to Azure and execute the scripts from "Step 1" below)
b) Docker (Containerizing the application and executing locally)
c) Install "Replace Tokens" Version: 3.6.0 by "Guillaume Rouchon" extension under AzureDevOps

NOTE: For simplicity; the below resource names reflect as they are in my environment; please rename as per your environment

#Note: Logon to your Azure Portal
Login-AzAccount

$subscriptionId = (Get-AzSubscription -SubscriptionName Polyglot-Development-PayAsYouGo).Id
$resourceGroupName = "pgt-seedwebapi-aue-rg"
$aksClusterName = "kube-seedwebapi-cluster"
$sku = "Standard_A2"
$region = "Australia East"
$kubernetesVersion = "1.16.10"

Set-AzContext -SubscriptionId $subscriptionId
New-AzResourceGroup -Name $resourceGroupName -Location $region

#NOTE: Switched to az commands as Powershell kubernetes "Az" commands are not supported in Australia East Region
# Using Kubernetes 1.16.10 version in below command; which is important; since this version defines how the kube-deploy.yml tempate 
# should be structured.
az aks create -g $resourceGroupName -n $aksClusterName --node-count 1 --node-vm-size $sku --kubernetes-version $kubernetesVersion

# Connect to the created Kubernetes Cluster
az aks get-credentials -n $aksClusterName -g $resourceGroupName

# Make sure the Kubernetes Cluster is indeed created and you are getting the desired number of nodes
kubectl get nodes

AzureDevOps pipeline azure-pipelines.yml defined in this project uses following three service connections

a) vramanavar1 : Connection to GitHub Code Repository
b) vramanavarDockerRegistry: Connection to Docker Hub Registry
c) kubeConnection: Connection to Kubernetes Cluster

NOTE: Point the above three connections to your sources (Code Repository, Docker Registry and Kubernetes Cluster). If you rename the service connection; 
      then have the same names used in azure-pipelines.yml file.

Pipeline is structured as below

1. Uses "ubuntu-16.04" Hosted-Agent 

2. Points to origin "master" of vramanavar1/seed-webapi project

3. Docker@2: This step uses multi-stage build Dockerfile. The image is built and pushed to DockerHub Registry via the service connection

4. replacetokens@3: This is extension added to AzureDevOps Organization. It is used to replace the tag of the Docker Image present in kubernetes 
                    manifest yml (SeedWebApi/kube-deploy.yml) file. Build Number is used to tag to newly created docker images.

5. KubernetesManifest@0: This steps connects to Kubernetes via the "kubeConnection" service connection. By now SeedWebApi/kube-deploy.yml would be pointing to
                    newly created Docker Image (Essentially, by Docker Image tag getting replaced by token replacement task in previous step)

Manually kicks-off the build job for the first time; which deploys the code to Kubernetes Cluster

Use the following commandlet to fetch the hosted public ip

kubectl get svc label-seedwebapi -w 

NOTE: Use "External-IP" column's IP from the output of previous commandlet. Use Postman and call http://<<External-IP>>/api/values. This should return 200 OK response

Perhaps, add some more values to api/values endpoint and push the code to the repo. 
This action should automatically trigger the build pipeline. 
Once the build succeeds. Hit the http://<<External-IP>>/api/values endpoint again; response should return new values (200 OK Response)

# Finally, when you are done working with Kubernetes commandlets; you can delete the resource group as below

Remove-AzResourceGroup -Name $resourceGroupName

Ideally for production workloads

a) Containers should start as Non-Root User
b) Privilege escalation should be disabled
c) ReadOnly access on the Root FileSystem