Intact GmbH | Parkring 6 | 8403 Lebring | Austria |

No Need For Magic—Your Data Science Dream Team

No Need for Magic—Your Data Science Dream Team

As with so many other things in life—working with data is a team effort. You need skills and knowledge about data storage, data extraction, data wrangling, statistics, mathematics, visualization and last but not least, in depth knowledge of the domain you are working in. When you read about artificial intelligence, predictive analytics and the promises they hold, the analytics industry is quick with offering you the silver bullet: analytical programs and platforms as self service offerings. They make you think that with the click of a button all your problems are solved and you get all the answers you are looking for without breaking a sweat. Some even promise, that you just have to type your question and you get the answer right away out of your terabytes of data.

We will eventually come to this point, but we are for sure years away from that, as experience and the latest developments have shown. And even if one day we have reached a point where we simply have to formulate the question that concerns us and the answer is delivered from the data without problems and satisfactorily, one thing must not be forgotten—there is a whole team behind every answer: Your data science team.

So Who Is Part of This Team?

As we all know, there is no „I“ in team, but let us take a closer look at the line-up of our dream team and see what it is that they are doing to extract the right information from your data.

Data Engineer

A data engineer has profound knowledge on how to organize, extract and wrangle data to ensure data operability. In the old days, data was stored on punch cards, then magnetic discs and optical discs, just to name a few. The larger the amount of data became, the larger was the need to structure it and make it more accessible and usable. The breakthrough for databases came in the 1960s and with it the famous structured query language (SQL) was developed. Today, also unstructured No-SQL databases and graph-databases with highly complex database models and high-throughput computing database models are in use. A data engineer has a broad skillset like data architecture, database solutions, data warehousing, hadoop-based analytics, coding and machine learning.

He is the one that builds the foundation of your analytics efforts.

Data Scientist

When the Data Engineers have completed their work, they hand it over to the Data Scientists. They are the ones who are digging into the numbers to draw often hidden insights for the business out of the data. Traditionally, data scientists came from fields like mathematics and statistics, since you need a very good understanding of these subjects to make sense of numbers and achieve certain goals with them like modeling data to predict audit non-conformities or find anomalies in audits. Data Scientists have skills in math and statistics, machine learning and they code mostly in R and python. They also have knowledge in data engineering, since they two work together closely with the data engineers.

Business Analyst / Executive Data Scientist

Data Scientists and Data Analysts are often mixed up and can also be one and the same person, if the data scientist has a profound knowledge of the business the company is operating in. A business analyst works with data to apply insight to the business strategy in general. A Business Analyst has good knowledge of the tools and pitfalls of data science and understands the data and needs of your company. He or she is the one that initiates the development of dashboards, predictive analytics and decision management tools. The skillset ranges from mathematics, statistics all the way to machine learning but not as profound as it is the case with a „real“ data scientist.

What a Business Analyst, who is often referred to as an Executive Data Scientist, also brings to the table are strong communication, presentation and management skills, which are essential skills in every team when it comes to communicate your findings.

Software Developer

Last but not least, you need someone in your team who takes your data and the developed data-models and transform them into some sort of source code to build applications that can be used by not so tech-savy people. The software developer in the data-science team himself has a good understanding of data science. That helps building meaningful applications and to communicate the needs throughout the team.

Your Data Science Dream Team

Your Data Science Dream Team

How Does The Team Work?

Now that we know the members of the data science team, we can take a closer look at what the workflow is like. Usually it is a pretty much a straight forward process with little to no changes in its flow. Instead of writing it down in some sort of essay, which might get a little confusing, I decided to just give you the bullet points in a clear form.

  1. Understand the problem(s) and need(s) – Business Analyst
  2. Ingest Data – Data Engineer
  3. Explore and understand data – Data Scientist
  4. Transform and clean data – Data Engineer
  5. Evaluate – Data Scientist and Business Analyst
  6. Create and build model or visualization – Data Scientist and Business Analyst 
  7. Communicate results – Business Analyst
  8. Deliver and deploy model / Visualization – Software Developer

After reading this you might think that it must be really costly to have a data science team in your company. Well, you are right! That is why most Small and Mid-sized Enterprises (SMEs) outsource data science teams. All of these positions are highly trained people with the need to be further trained on a constant basis. There are of course companies who can do this incredible specialized work to help you get the best out of your data. One of them is Intact’s Analytics team which I am a part of myself. You can find out more by clicking the button below. We are looking forward to hearing from you!

Find out More


Write a comment

Deine E-Mail-Adresse wird nicht veröffentlicht. Erforderliche Felder sind mit * markiert.