Recruiting Data ScientistsThe Challenges of Hiring and Vetting Data Professionals
Data and analytics are no longer a luxury; they’re the proven way of doing business. In fact, any company that’s not using available data might not be able to dominate the market.
For competitive reasons, hiring managers would need to distinguish so they can hire the right people for the right positions. Besides, we’ll discuss the must-have skills of good data professionals. Next, we’ll touch upon vetting candidates in the selection process for those who possess the best technical and soft skills.
Data Scientists, Data Analysts, and Data Engineers
Data scientists and data analysts are two of the most lucrative jobs in IT, which also have impressive salaries. While they have some similarities, there are fundamental differences, which a hiring manager must distinguish to recruit the most suitable individuals for the positions. In addition, there are data engineers whose task is bridging both the scientists and the analysts.
In a nutshell, data analysts are experts in examining and analyzing data sets to identify trends, create charts, and other visual presentations for decision makers. In other words, they’re “the data readers.”
Those data scientists, on the other hand, develop and deploy processes for various data modeling and data collection with algorithms, predictive models, analyses, and prototypes. In other words, they’re “the thinker behind the data collection and processing.”
Scientists who work on data can be categorized into Type A (Analytic) and Type B (Builder). The Type A data scientists are known for being more analytical in their work and might not be dealing with coding. Type B data scientists focus on developing data-driven products, which often require programming.
Math skills and other technical skills are crucial for data scientists while being a data analyst requires more analytical and the ability to connect the dots. This being said, the skills of these two jobs can overlap, but each has its own distinctive specializations.
Data engineers, however, work with both data analysts and data scientists. Their primary tasks are making sure that data sets are accessible in databases. They’re known for integrating models into production. Thus, they’re more likely to be more involved with coding and programming.
Must-Have Data Science Skills
For a data scientist, it’s advised that the candidate has a Master’s or a Ph.D. degree in Mathematics and Statistics, Computer Science or Engineering. The preferred technical skills to have include the following.
Tools: NumPy, SciPy, Pandas, Scikit-learn, NLTK, IPython, Matplotlib, Git, MySQL, PostGIS.
Techniques: Linear Regression, Logistic Regression, KNN, Random Forest, Gradient Boosted Trees, SVM, Clustering, PCA, Neural Network, Natural Language Processing (NLP), Collaborative Filtering.
More advanced data scientists are likely to have in-depth experiences in one or more of these skills: Deep NLP, Text Mining, Tagging, Syntactic Parsing, Sentiment Analysis, Contextual Text Mining, Epsilon-Greedy, Softmax, UCBI, Dynamic Programming, Monte Carlo, Temporal Differences, TD Lambda, SARSA, Q-Learning, Deep Learning in Keras and TensorFlow, AI, Artificial Neural Networks, Figaro, Scala, Java, .NET, MapReduce, Spark, SVM-Light, WEKA/MEKA, iPython, Jupyter, OpenAI Gym, NetLogo, and Cloud Computing Google App Engine and Amazon EC2.
At the same token, data engineers are likely to use these primary tools: Python, Java, Scala, SQL, Spark, Airflow/Jenkins, Cloud Computing, and Docker.
Vetting Process for Selecting Data Professionals that Fit Your Business Needs
Each organization is unique, so is your business. The first and foremost thing to do is discerning your business’ requirements for data. In other words, you’d need to know beforehand the tasks that those data scientists, analysts, or engineers would actually do, once you’ve hired them.
Since every individual has a different combination of skill sets, searching for the most suitable candidate isn’t that straightforward, unfortunately. For this purpose, it’s recommended that you create a list of what your organization would require.
For instance, the types of data to collect and what the purposes are. Also, the Key Performance Indicators (KPIs) used to measure each candidate’s performance would require them actually to perform well on the tasks.
This would require the hiring manager to truly understand the tasks to execute, the measurement metrics, and how to select the most suitable candidates for the position. In other words, he or she must have sufficient knowledge to understand both sides of the fence.