Market for Data Scientists
A lot of talk is made of the difficulty in finding the right people needed to kick-start data science initiatives. However in my experience, the most fundamental roadblock is the unstable foundation of the data estate in most organization. Only when quality data is obtained in sufficient quantity can data scientits do their magic.
The minimum expertise data scientists need is a combination of three critical skills: deep business knowledge of the domain area, understanding of the data feeds and formats in relevant sources, and the algorithms and statistical techniques not only to create predictive models but also to evaluate their characteristics, impacts, and applicability in production environments.
Hiring data scientist out of school can be a good option, but keep in mind that junior data scientists will be week on the business side and you should expect much longer ramp up time. Another common question is whether existing employees be retrained to become data scientists. Business analysts with technical inclinations are well positioned to become data scientists by learning the statistics, algorithms and processes of data science, but the learning curve is steep. Recommended courses exists on edX and LinkedIn Learning.
You might be wondering, where are data scientists hiding? The research I did on LinkedIn, Glassdoor and Indeed can give a glimpse of the discrepancy between US and Canada as well as between Tech leaders and the #2 industry of financial services.
If data scientist are difficult to find, and quite picky in where they want to work, another approach is to leverage the vast ecosystem of partners that we have locally to work on your data science initiative.
Another approach is for your developers to infuse your app and business processes with pre-built AI services: Azure Cognitive Services. This can be a quick way to get started for more generic AI scenarios.
As an example, consider how Uber decided to add face recognition technology in their app by leveraging these cognitive services in order to increase the safety of passengers. No data scientist required for this!
Here's a list of helpful resources below:
Regional Sales Manager, Mid-Market EMEA @Datadog
6 年Totally agreed! A lot of companies think hiring data-scientists will magically kick-start an AI strategy, and overlook the importance of quality data and the required infrastructure for data-scientists to be productive. Oh and you missed mnubo as a local Microsoft data-science partner ;)