Companies that provide insight and analysis are striving to form a perfect data science team for themselves. There are usually two ways to go.
Most companies are struggling to choose the first way: looking for these very expensive and rare unicorn talents, that is, independent individuals with these skills at the same time.
The perfect data scientist has mastered mathematics, statistics, programming and communication technology perfectly. These people not only have the professional technical ability to complete complex work, but also can explain the possible impact of these technical work to non-technical personnel.
Apart from the fact that these data scientists will be expensive, it is impossible for these talented scientists to work 24 hours a day, 7 days a week.
Of course, this is not the only way.
Aristotle, one of the earliest scientists in human history, once said: "The whole is greater than the sum of its parts", which gives us some enlightenment. Instead of looking for these popular candidates with three skills at the same time, it is better to choose people with one of them as a team. After all, no one can always solve the organization's growing demand for data science research. It needs a mathematician to be in charge of in-depth research, and a person with interdisciplinary knowledge to integrate horizontally, and finally form a perfect team.
◆◆◆◆◆◆ The vitality of the data science team
The ultimate goal of any data science team is to become a problem-solving machine, a team that can constantly stir up value in a changing environment. More and more abundant data makes it possible to answer once unanswerable business questions, which raises customers' expectations of insight into complexity to a new height. However, there is no mature methodology and solution for this chain reaction. As the input becomes more and more diversified, the matching skills needed need to be diversified. None of the three characteristics of the "cool nerd" team is indispensable, so the collective wisdom of this team is really the driving force of today's data world.
Obviously, no two pieces in the team of perfect data scientists can operate independently of the third piece. In addition, mining and maintaining the internal balance of the team of data scientists can bring the greatest accuracy and relevance.
A mathematician/statistician
After understanding the conditions needed for relevant theories and results, these trained scholars established advanced models based on these inputs.
Program arranger
This personal architect is responsible for cleaning, managing and trimming data, and building simulators or other high-tech tools to make data more convenient and easy to use.
Communicator/content expert
Experts who turn technology into business, based on past knowledge, use their own overall view to help find the connection point between technology and user needs.
The mutual support of these skills makes the team complete and has perfect data delivery ability:
The work of mathematicians/statisticians depends largely on programmers. The concept of "garbage goes in, garbage comes out" is very applicable here, that is, it is difficult for scientists to build useful models if programmers do not clearly acquire and manage data. In addition, mathematicians and programmers rely on the knowledge of communicators. Even if the data is perfect, it is consistent with the statistical conclusion. None of this is meaningful if it cannot be directly related to the business problem to be solved. In addition, the team with internal imbalance will face some difficulties of insufficient preparation and unable to deliver perfect works.
◆◆◆◆◆◆ Is it purchased or self-built?
Today's world is full of high-speed data, and enterprises are faced with a choice. Traditional programmers who write code editing questionnaires and collect data are integrated into insightful organizations. However, many of them have no formal training in mathematics or statistics. Similarly, those business talents who are customer-oriented and have numbers and quantity minds should also occupy a place in team building. It is feasible to train existing mathematical or statistical talents, but the long process requires great patience. If organizations recognize and believe in their talents and choose this path to form a team, it also points out the gap that needs to be filled to create a perfect team.
Organizations have long known the value of data, but without people's participation, no matter how large the amount of data and how deep the details are, it is difficult to realize the valuation of data science of $30 billion before 20 19. Interpreting, filtering and correcting all kinds of data by a balanced team will accelerate this growth and enhance the importance of data science.
Many people think that Hillary's concept of "cool nerd" only applies to individuals. But in fact, we must realize that the collective "calmness" of the team is also full of potential.
When an organization is building and recruiting a data science team, perhaps the purpose of the team can be simply called "If you can find nerds, keep them." . But if you lack a team that brings together all kinds of unicorn talents, create one. "