Companies are increasingly data-driven, which has led to the rise in several different types of data-based disciplines. Indeed, the demand for people with expertise in data is so great that there are global shortages in all data roles. But what are the different data roles, and what do they do within the organisation? If you’re looking to build your career in these areas, just what do you need to know?

## Data Science Data science is the all-encompassing field of expertise for people that work with data. In simple terms, it refers to those experts that take the data that comes into the organisation from a variety of sources (most of them online and digital) and works with it to assist business decision making, strategic planning and so on.

Data science has a role in just about everything to do with modern businesses. Data scientists will identify new business opportunities because the data will tell them customer trends and the movement of the markets. Internally, data scientists will help the business run more efficiently and effectively. Data scientists will also help the marketing team more effectively reach customers, the IT team more effectively secure the environment, and the logistics team deliver products more quickly and cost effectively.

Data scientists are often related to the IT team within an organisation, and will be involved in coding, but at the same time perform separate functions and are not considered IT professionals. Their primary fields of expertise are software programming, but more importantly, statistics and mathematics.

Data scientists will often specialise in a particular vertical, too. For example, a data scientist in the medical industry might assist their company with diagnosis and treatment, while a data scientist in finance might be focused on developing models to improve trading or fraud detection. It is a broad field, and can be a lucrative one for anyone that develops the right skillset.

## Data Analytics Within data science, one of the most common applications is data analytics – in fact, most data scientists will spend a big part of their careers involved in analytics. In simple terms, data analytics involves drawing insights out of data, and this can take four different forms: 1) Descriptive analytics. This is valuable to organisations because it helps to track performance over time. The data scientist takes large data sets and uses them to track successes or failures for activities, and descriptive analytics are used to form KPIs and describe the expected outcome for activities. 2) Diagnostic analytics. This form of analytics explains what has happened. It involves taking the findings from descriptive analytics and then digging deeper to explain the root cause and trends for why something is happening. 3) Predictive analytics. While the previous two forms of analytics were looking back, predictive analytics look forward to explain what will happen. Here the data scientists will take historical data and analyse trends to understand what will likely happen into the future. 4) Prescriptive analytics. Finally, prescriptive analytics will explain what an organisation should do when faced with a decision. This again involves looking at historical data, but rather than using it to make predictions, prescriptive analytics is focused on the current moment and delivering outcomes.

A data scientist will typically be building models for all four types of analytics within their organisation, because all provide valuable insights that can help the organisation understand how to better shape the business and where its best opportunities lie into the future.

## Machine Learning Machine learning is a practical application of data science in that it is used to build tools that assist the organisation now. Essentially, machine learning is a precursor to AI; a data scientist will build a model or application, and by feeding it a lot of data, the model will start to “learn” expected outcomes. The AI layer then sits over the top and allows the model to take actions without further human involvement.

A good and easy example of machine learning in action is IT security. A machine learning application will constantly scan an IT environment and, when it detects an anomaly (because it has been fed so much data on what is and isn’t unusual activity), it will immediately raise a red flag for the IT team to investigate. Machine learning is essential to modern business precisely because it allows for a “hands off” approach to many activities within the organisation, and so data scientists will find a lot of their time is occupied building machine learning models.

Data science is such a broad field that there are many ways to specialise within it. The foundations are in understanding maths and statistics – not technology, as is often assumed – and applying that knowledge in a way that helps an organisation draw insights into what is going on, internally and externally. Professionals that can develop expertise in data science find themselves well placed, because this is a skillset that is very much in demand and businesses are willing to invest heavily in having it.