Honing Your Data Translator Skills

Scaling AI Catie Grasso

If you spend a long enough time cleaning and preparing a dataset, chances are you will grow to understand its nuances and behaviors. But once you’ve put in all of that work, how do you communicate the important information to someone who doesn’t have the proper context or the time to familiarize themselves with the nitty gritty details, such as a business stakeholder? How can they possibly ever understand the thoroughness of your analysis and the steps you took if they weren’t involved (and it wasn’t their job function to be)?

Enter: data translators. Many data leaders are beginning to embed them into the organization to effectively translate business needs into data needs and communicate valuable insights to the relevant business teams. 

Gartner refers to them as business translators and states, “this could be a business-savvy data scientist or citizen data scientist, an analytically-minded business person or process engineer (process modelers or business analysts focused on process design) who is mindful of business optimization opportunities derived from analytical assets.”* Here are a few ways you can hone your data translator skills to help infuse more agility into the transformation and delivery required of large-scale data projects:

  1. Keep a pulse on the latest trends in business analytics (and its application across data science and machine learning), as you are the one who ultimately facilitates the process of using the outputs from the data science team to tackle business problems.
  2. Be a team leader and strong communicator, staying up to speed with business acumen and domain expertise. You’re a “common ground” between business and data teams and you need to know how to collaborate in an inclusive way.

two colleagues at a whiteboard3. Move out of spreadsheets for data prep and analysis and into an all-in-one platform. Doing so will remove inefficiencies across the data-to-insights pipeline, both from a manpower and actual output perspective. Further, in Dataiku, for example, the visual flow enables clarity for anyone consuming an analysis they didn’t make themselves — it’s digestible and easy to follow.

Last, even though your role requires general technical proficiency and doesn’t go deep in the weeds, doing this work in a tool that allows for machine learning will mean you can move to more advanced projects in the future. When you do, those projects can be more easily expanded and developed if you don’t need to toggle back and forth between a spreadsheet and machine learning platform. 

→ Get Tips on How to Transition Out of Spreadsheets for Data Prep

4. Know what to prioritize instead of just digging for potential use cases. It’s your job to identify existing business problems or processes that can be streamlined and bring those to the data team after gaining buy-in from stakeholders. Of course, though, your industry and company business objectives will greatly influence how you structure this prioritization process.  

Data translators (who, over time, evolve into trusted figureheads across data teams) can be injected into the appropriate pockets of the business to identify data requirements, oversee data work streams, and act as mediators and and go-to points of contact for members of the development and operationalization teams, as well as for executive stakeholders. 

*Gartner - Use 3 MLOps Organizational Practices to Successfully Deliver Machine Learning Results, 2 July 2020. Shubhangi Vashisth, Erick Brethenoux, Farhan Choudhary

You May Also Like

Talking AI Democratization With Dr. Anastassia Lauterbach

Read More

6 Top-of-Mind Topics About AI & Trust in 2024

Read More

3 Concrete Ways to Drive AI ROI

Read More

Alteryx to Dataiku: The Visual Flow

Read More