Posted

This post is contributed by Naomi Keller and Caroline Evans, Burtch Works’ data science and data engineering recruiting specialists.

Five years ago the data science hiring market was awash with unicorn headlines, but, as we reported earlier this year, we’re noticing a distinct trend towards specialists instead of generalists. And while many data scientists are still equipped to work on every stage of the analytics lifecycle (including data acquisition, transformation, analytics, etc.), many are now working on teams with other analytics professionals and data engineers rather than covering every step in the process on their own.

This has led to more diversification in data scientist teams as companies look to drill down on specific use cases. Although there is still plenty of overlap – data scientist hires are still expected to have strong technical skills such as experience with Python and machine learning algorithms –we’ve noticed four areas where employers are seeking slightly more specialized experience.

1. Natural Language Processing (NLP)

NLP is the application of artificial intelligence to blocks of human language, and often used to analyze or interpret unstructured text. It has been used for processes like sentiment analysis, which involves categorizing opinions from text and is commonly used to examine customer reviews, but more companies are looking for NLP experts to help them understand even more about their text data.

In healthcare, where lengthy patient records and doctors notes abound, NLP can be used to evaluate insurance claims or to better understand patient outcomes, which we’ll cover further later in this post. NLP is also put to use in call centers to more efficiently serve customers or with virtual assistants like Alexa to better understand user needs.

Common skills/tools: Deep Learning, Python, Machine Learning, Tensorflow, Computer Science, Computational Linguistics, Neural Networks

 

2. Computer Vision & Image Processing

Computer vision and image processing involve AI systems that are tasked with analyzing visual information, including from static images as well as videos. Data scientists with this specialty might be creating systems for insurance companies to investigate car accident photos, or even working for a beauty company to analyze photos of a customer’s face to provide better product suggestions.

Autonomous vehicles rely on advanced computer vision to see the world around them, and we’re seeing more opportunities related to image processing as companies seek to use neural networks to automate processes that could previously only be done by an insurance agent, beauty consultant, or even the driver of a car.

Common skills/tools: Deep Learning, Machine Learning, GPU/CPU Programming, Object Detection, 3D Reconstruction, Pattern Recognition, C++, OpenCV, Matlab, Python, Spark

 

3. Ad Tech

Data science applications in advertising often fall under the generic umbrella of “ad tech”, which can include programmatic platforms that manage real-time bidding for ads, as opposed to the traditional process of manual negotiation and orders by humans.

Although digital natives like Facebook (and many others) have had these systems in place for quite some time, we’re seeing ad tech experience become a common ask among an increasing number of legacy organizations as well. Some firms seek to bring these capabilities in-house and create their own data science teams devoted to advertising applications for more control and efficiency, rather than relying on an outside marketing agency.

Common skills/tools: Python, Machine Learning, Hadoop, Programmatic Advertising Marketplace, Real Time Bidding

 

4. Internet of Things (IoT)

The Internet of Things, which includes internet-connected devices, and other sensor-driven processes, has also proven to be a lucrative area for data scientists to specialize in. The use of sensors (and the data they produce) has exploded over the past several years, and industries ranging from insurance to agriculture want in.

IoT data scientists can be by companies tracking their employee’s movements with RFID tags to better understand worker allocation and productivity or by retailers looking to more effectively measure inventory and manage their supply chains. This is not to mention the already booming areas of predictive maintenance for industries like manufacturing and transportation and developments of IoT in operations research.

Common skills/tools: Python, Machine Learning, Sensor Data, Telematics, Predictive Maintenance

 

Emerging Industry Spotlight: Healthcare

As we mentioned before, healthcare is an industry with vast troves of data, and is also an industry where data scientists can find opportunities with a wide range of responsibilities. With copious amounts of messy data, new techniques, and an increasing number of data sources (including everything from patient records to x-ray images to fitness tracker data), the opportunities for data scientists with healthcare experience are substantial and growing.

 

As the data science field continues to mature, it will be interesting to see how the specializations vs. unicorn data scientist segments evolve. Improvements in technology and increasing automation will likely have an effect on these divides as well, but only time will tell!

 

Interested in our salary research on data scientists? Download our studies using the button below.

Click to download our free salary reports
 
We hope you found this information interesting, and if you’re looking for opportunities or to hire professionals in data science or data engineering, be sure to connect with us (Naomi Keller and Caroline Evans) on LinkedIn!

3 Responses to “Beyond the Unicorn: 4 Developing Data Scientist Career Paths”

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.