Table of Contents

Recipes in action

The biggest problem AWS has, compared to Azure that is overtaking them in pure mass, is all their services and videos are developer-focused. The elaborate network of business partners and by leveraging the office365 offering makes it hard for AWS to compete. If we are able to combine snapshots of CF or CDK templates to simple business requirements, I think we are able to create scale-able data products.

Apart from its terrible generic, nonsense naming, AWS Solutions, is what we need to build as a recipe. AWS solutions, hence our recipes come in one of these four flavors:

  1. Reference Architectures: e.g. still there is a gap in the market for a ref architecture of Data Mesh, although unfortunately, AWS took this category a bit light by limiting it to mostly diagrams and no code.
  2. 🔥 Constructs: this is a relatively new flavor but being promoted by AWS as they promote CDK further and further. Here if we move fast we'll have an early mover advantage. Train yourself on CDK!
  3. Implementations: This has been the classic form of AWS solutions. Data Mesh, SageMaker pipelines done right, etc can all be formulated as an implementation. In practice, there hasn't been anything but CloudFormation templates and stacksets.
  4. Consulting Offer: here things are a bit mixed between promoting our consulting capabilities and templated/scripted solutions.

<aside> 📌 Interested in diving deep in CDK? Check out CDKPatterns maintained by Matt Coulter.


Characteristics of a Good Recipe