What exactly is a ... Data Scientist?

Posted by Prof. Dr. Nick Gehrke on Jan 27, 2017 5:00:00 PM

... and what has this to do with auditing?

The job profile of the Data Scientist is still young, but is often searched for on the job market. They are required in many industries, such as:

• Banking and insurance 
• Trading
• Business and organizational consultancies, market researching
• Social Media, Telecommunications, online tradinging and network management
• Bio-, pharmaceutical, chemical and medical industries
• Logistics




In 2012, Tom Davenport, Professor at the Harvard Business School, has described the competence profile as following: „… a hybrid of data hacker, analyst, communicator, and trusted adviser. The combination is extremely powerful – and rare.“

In times of "big data", Data Scientists are experts in demand, who are paid above average and enjoy great freedom in companies as "gold diggers". Using methods of mathematics, computer science and statistics, they gain facts and knowledge from large amounts of data, the "gold of the 21st century", and discover new business areas. In addition, they are something like interpreters. They formulate the data records into legible results and display the essential information in a comprehensible language.

Data Scientists are trained in statistics, graph theory and other mathematical fields, and are proficient in methods such as data mining, process mining, machine learning and natural language processing (NLP). Added to this is knowledge from practical computer science. Knowledge of operating systems, databases, networks and data integration tools, as well as the most important programming languages and analytics tools are mandatory. Furthermore, knowledge about the Hadoop ecosystem, social networks and other systems from the internet and big data environment is a compulsory requirement for professional practice. The competency profile is that of an all-round talent and accordingly (currently) difficult to find.


The Data Scientist and the financial function within the company

The question whether a controller can assume the tasks of a Data Scientist must be clearly denied in the context of the described competence profile. The current opinion in the industry is, that it is illusory to believe that controllers could also assume the tasks of a Data Scientist. However, controllers should know the job profile of a Data Scientist as well as the possibilities and limitations of Big Data. The cooperation between the tasks of a controller and a Data Scientist is an important source for the future economic success of companies.


The Data Scientist and Auditing

The advancing digitization also places new challenges on internal auditing in the selection of the audit methodology. Data Science offers the possibility to consider the analytics of data masses as a test step within an audit and in this way to create an additional benefit. This means, however, that the internal audit department must also acquire expertise in data science in addition to the already acquired competences, such as finance, business management and compliance. Since an individual auditor can hardly have all the competences mentioned above, these should be at least available within the team. If necessary, remember to include an external Data Scientist.

Along the lines of internal auditing, the external auditing is placed before conditions that were changed by digitization: the flood of data, the appropriate audit methods as well as the concern of finding young recruits within the auditors underline the need for efficiency gains. The surge in job advertisements for data scientists in audit centers, as well as first attempts to use artificial intelligence in this area, underscores this.


zapliance helps with the digitization of your auditing

The digitization of auditing can only be achieved with a modern data and process analytics software. Download zapliance for free or contact us if you have further questions.



Topics: Data Analytics, digitization, Data Mining, Process Mining, Data Science, process analytics

Blog Commentary