Data architects need an elevator pitch in their back pocket on the who, what, where, when, where and how of data architects and data modeling. We work in a discipline that many people do not understand. Here’s my speech as free of technical jargon and acronyms as I can be.
The data model organizes like data into building blocks and shows the relationships of these blocks to each other. This picture guides how the database is built and applications access this data. Think of it as the database’s blueprint.
Data modeling is the collection of techniques used to draw this blueprint. Think of it as the architectural principles and building codes that assure the database foundation is up to par and meets the demands placed on it.
If you want to understand data and how it interacts with other data, it is a good idea to create a data model. People relate to visual presentations. The data model simplifies the technical view of data into one that is more easily understood by the non-technical individual.
It might appear that a data model is not necessary to design a database. Would you build a house or automobile without a blueprint and detailed specifications? The answer is obviously “No”. Investing time and resources in data modeling results in a better product in the end.
Data modeling is part of the IT application development lifecycle. It is rooted in the majority of IT projects requiring database structures to hold and manage data. Data modeling can also be found other initiatives where the organization and rules of data need to be understood.
The data architect, also known as the data modeler, creates the data model. They must understand what the database must accomplish and the data that supports those needs. In addition, the characteristics of this data and the business rules defining how data flows must also be captured.
Data architects work closely with both technical and non-technical folk to do their job. They look at data from an enterprise perspective and ideally look to integrate data across the business.
Data architects use specific software to create data models. Data modeling tools enforce methods and standards that help the data architect deliver a high quality data model. These tools also capture the characteristics of the data, known as metadata, which helps everyone understand what the data is and how to use the data.
Data modeling is an iterative process. It starts with capturing data requirements and progresses through the creation of the database. In traditional data design, data modeling starts with a high level blueprint of the basic pieces of data. As requirements are discovered, the data model shows how the database will work from a non-technical perspective. It is transformed into a more technical data model optimized to take advantage of database and IT technologies. The end result is a database where the data is stored and retrieved.
Faster, shorter development methods, commonly known as agile, are becoming common. Traditional data modeling processes are chunked into smaller pieces delivering subsets of the data model in a tighter timeframe. Componentized agile delivery more tightly integrates data modeling with software development teams.