Both types of engineer must be passionate learners since technology keeps evolving, according to an article on the mckinsey.com website.
Machine-learning engineers
These software engineers know how to play with data and are able to work in scalable computing environments (Cloud, Hadoop, etc.). They also can adjust the algorithms when necessary and apply machine-learning algorithms because they:
- are able to choose the proper model and select learning procedures which fit the available data
- know which trade-offs are connected with various approaches
- grasp how different parameters affect machine learning
The data used are often customer data (sources include social media or purchase activities). Machine-learning algorithms can help here. Evaluation of data sets and identification of patterns and trends helps to get the right data for such algorithms.
DevOp (the integration of development and operations) engineers
DevOps is a term composed of the English words development and operations. It represents the way of work that integrates developers with operation teams.
Managing digital resources: this role has been created with the rise of a cloud infrastructure computing ecosystem. Infrastructure resources (networks, servers, storage, applications and other services) are often provided externally. Companies therefore hire DevOps engineers who are able to navigate in this field.
These workers create tools for developers, who then have infrastructure resources available at the click of a button. Traditional processes were lengthy and complicated.
They should be able to use infrastructure-automation technologies (Chef, Puppet, and others), cloud platforms (AWS, Azure, and others) and a few more things besides. They must be partners to their colleagues from business so they can help them solve problems as they arise.
-jk-