Home / China / Innovation

Tips for recruiting a top data science team

By Yan Dongjie | chinadaily.com.cn | Updated: 2017-07-12 11:20

Tsinghua University Institute for Data Science and Big Data Digest released the first Roadmap for Building a Top Data Science Team on Tuesday.

In the era of big data, data science teams, as the core player in a data-driven enterprise, have been attracting increasing attention in many industries. However, for enterprises that target data-driven operations, how to build an effective and well-coordinated data science team still remains a question.

The university's roadmap provides some answers to questions such as: Does an enterprise need an independent data science team? When and how should the enterprise build the data science team? How to measure the value brought by the data science team?

"The data industry is at the starting period. The report examines the status of the field, points out the existing problems, and tries to give solutions, which is practical and instructive for not only the industry but also the trainings in schools," Han Yishun, the executive deputy head of Tsinghua University Institute for Data Science, said.

Wang Decheng, the founder of Big Data Digest and the main initiator of the project, released the roadmap on Tuesday.

Wang said that the giants in the finance and IT industries currently lead the contest of data science team building. The high informatization achieved in their early development stage gives those enterprises considerable advantages over companies in other industries. Among them, enterprises in the finance industry have the highest rate of outsourcing data-related operations, and adopt an "outsourcing + endogenous growth" strategy.

On the contrary, companies in the IT industry have data science teams that are more centralized, outsource fewer data-related operations, and are more likely to be independent of other teams. Following companies in the finance and IT industries are enterprises in the transportation, healthcare, public administration, energy and education industries, while companies in the accommodation, catering and agriculture industries are more or less still warming up at the scratch line.

"More than half of data science teams have reported shortages of data science talents," Wang said.

The roadmap is a collaborative project by Tsinghua University Institute for Data Science, Big Data Digest and TsingData Institute.

This three-month project involved analysis of more than 50,000 entries of worldwide online data, thousands of survey responses, and interviews with the leaders of 10 top data science teams.

The Roadmap points out that in most cases, organizations and institutions have a fixed budget for talent recruitment. Therefore, finding properly qualified talents under budget constraints becomes the primary problem that concerns the leaders of those enterprises.

Among all data-related positions, NLP engineers, data scientists, machine learning engineers and algorithm engineers are the highest paying. The building of a professional data science team requires intensive search and cultivation of talents, an optimized talent management structure, normal team operations and the stable long-term development of the enterprise.

 

Editor's picks