SI4 : Promise of AI ~ Hire Machine Learning Engineers not Data Scientists

There are two broad classes of machine learning professionals, academic (data scientists) and industry (machine learning engineers). To have successful outcomes in building AI applications, an organizations needs the latter who are practical orientated. Models need to move beyond the proof of concept stage and demostrate scalability while being stable. Deployment must also be flexible and easy such that it can handle the throughput speed.

It must be noted that while algorithms are a fundamental part of any machine learning solution, they are however small in terms of contributing to the entire process. The surrounding infrastructure matters when compared to the small black box in the diagram and as validated in the famous paper by Google : Hidden Technical Debt in Machine Learning Systems