Commercial Optimiser - is an app that determinnes the optimal allocation and calculates the achieved rating score for a set of TV commercials based on several criteria.

Default settings:

• 3 demographics (Women 25-30, Men 18-35 & Total People 18-40)
• 9 commercials (each commercial has a demographic it will target)
• 3 commercial breaks with ratings per demographic

Restrictions:

• Finance type commercials cannot go into Break 2
• Commercials of the same type cannot be next to each other within a break

The application supports all types of mobile devices and works well on desktops. The default commercials data is stored in the database and is loaded into the application at startup. The user is given the ability to enter new rating for each commercial for a corresponding brake.

• This application follows the modern serverless design and the best practices of building and hosting cloud based serverless public facing applications.
• The application consists of two main parts - the Reactjs UI app and the .NET Core Lambda api.
• This design achieves great economy as it makes unnecessary running a full-scale web server and this is also in line with all the principles of the microservices architecture.

• The UI app is built using ReactJs and Bootstrap 4. It is compiled and then copied over as a package of static files into AWS S3 bucket. The bucket is set up as a static file server. In theory the UI part can be hosted on any sevrer capable of serving static or dynamic applications. It does not require IIS, Kestrel, NodeJs or any other specific server. Once loaded into the browser the app will start making REST calls to the the backend service which is also hosted in AWS as a serverless Lambda. In theory the backend service can be swapped with another service capable of serving content over RESTful connections.
• The UI app supports mobile devices and desktops equally well.

• The api is based on the .NET Core 3.1 AWS Lambda template. The logic is implemented using C#.
• The api is run as a serverless application.
• The api uses the serverless database technology DynamoDb as it's main data storage provider
• All the default application data like ratrings, commercial ids etc are retrieved from DynamoDb at startup.
• The api uses two DynamoDb data tables for its operation. The tables are provisioned during the initial deployment of the api (procedure is decribed below).

• The app can be built and run locally

to start the app navigate to .\optimizer\app directory and run

npm run start

to build the app for deployment run

npm run build

now you can copy the files from the build folder to the server of your choice

to build and package the lambda project locally navigate to .\optimiser\api\src\optimiser directory and run

dotnet lambda package 

this command will zip the packag into \Optimiser-api\src\Optimiser\bin\Release\netcoreapp3.1\Optimiser.zip

to run tests navigate to .\Optimiser-api\test\Optimiser.Tests directory and run

 dotnet test

In order to use commands given below you'll need to replace [my-bucket-name] with the name of your own s3 bucket

Create an s3 bucket for the source code of your lambda by running following command from \optimiser\api\src\Optimiser directory:

aws s3 mb s3://[my-bucket-name] --region ap-southeast-2

To prepare the package for the deployment to CloudFormation and to transform the template.yml run:

aws cloudformation package --template-file ./template.yml --output-template-file sam-template.yml --s3-bucket optimiser-api

To deploy the packaged source code to aws lambda using CF run:

aws cloudformation deploy --template-file ./sam-template.yml --stack-name optimiser-api --capabilities CAPABILITY_IAM

initially the data tables are provisioned empty. In order to populate them with data the PUT api can be used, provided the url where lambda api has been deployed is known. This will populate the tables with default data defined in code.

PUT https://[lambda-hostname]/Prod/v1/default

Currently the app is deployed here here.