Topic

What players should be on a data science team? – roles and responsibilities.

Introduction


While conducting research on this topic I came across the following resource (2) which proposes three distinct roles essential in any data science team: Software Engineer, Data Engineer and Data Scientist. In many organizations, a Data Science Team Manager is also part of the team to help bridge some gaps between the team and other stakeholders. It is important to note that in many organizations (especially start ups), one person may serve various roles until the buisiness needs and data science team structure becomes more apparent. The data needs of a growing business may drastically change over time and therefore, data science teams (roles, responsabilities, etc), must evolve as well.


Roles & Responsabilities


The image above illustrates nicely the various skills and tools nessesary in a data science team. SOURCE


Data Science Team Manager/Lead


According to the Harvard Business Review 1 an effective Data Science Team Manager not only builds/maintains a diverse and high performing data science team, but they also show team members that their interests are valuable and can connect to team goals. An effective team manager has a clear understanding of the business goals behind projects, and is able to anchor the teams work to the context of the business’ broader organizational strategy. This often means ensuring team members are regularly invited to product and strategy meetings. DS Team Managers are also able to keep all memebers of the team productive by leveraging their strengths, giving ample time for feedback and debrief so as to keep everyone feeling like their contributions are valuable.


Data Engineer/Architect


Daily activities for a Data Engineer on a data science team include:

  • Developing and maintaining database architectures that support buisness requirements
  • Develop prcoesses for data modeling, mining and production
  • Employ multiple languages as needed to “marry” systems together, and reccomeend ways to improve data reliability, efficiency and quality


Data Scientist


Daily activities for a Data Scientist on a data science team include:

  • Answering industry and business questions using both internal and external data sources
  • Create sophisticated analytics programs, ML and statistical methods to use data for predictive modeling and forecasting
  • Use predictive analystics to automate models
  • Communicate data narrative based on analysis that is accessable to all stakeholders


Software Engineer


The main goal of the software engineer on a data science team is turn the data insights and other information stored in data warehouses into scalable products that can be used by multiple stakeholders. For example, it is often times the software engineers job to create dashboards and entire applications for customers and employees to have access to the data that they need in a convenient and practical way. With this said, the software engineer usually lends their expertise in user interface design, building APIs, fixing bugs on applications, examining, testing and deploying models and updating application functionality as new data becomes available. As with all other roles, this role is largly based on the needs of the business. Larger businesses often increase productivity by hiring software engineers who can create user firendly data interfaces for employees, so that they can make data informed decisions as data becomes available.


Other Roles

Depending on the needs of the business, Data science teams can also include a range of other more specific role. For more information on these roles see the following resource DataCamp Community Infographic. A roles and responsabilities for a data science team should be designed accoding to specific data needs by


Questions to Spark Discussion


  1. It was briefly addressed that effective Data Science Team Managers have both Management and Community Building skills. What sets of skills do you think is most important for a Data Science Team manager to have: Management, Community Building or Technical? Why?
  2. Do you agree with these specific roles? What role do you think was left out? Why is that role important in many data science teams?
  3. Which roles from the DataCamp Infographic do you think are needed in a data science team? Which ones do you think are not needed?

THANK YOU FOR READING!