For experienced professionals from non-software domains, which options is better between IoT and machine learning?


If there is confusion between IoT and machine learning, it is fair. After all, both are quite hyped and promising. For anyone to get a job in any of these two, it should "make sense" from the hiring manager perspective. They would ask about career history and check how this is the next logical step for you. This way, they are convinced that you are genuinely interested in the subject and also qualified.

Regarding IoT vs machine learning, let us first understend these two.

IoT is a broad 'solution architecture' of how devices are at the 'bottom' , then they communicate to some server somewhere, the analytics are run on the data and some control or display action is taken either in real time or offline. i.e. someone who works purely on an embedded box like hardware engineer, mechanical engineer etc could easily claim IoT skills. You could be a someone with mech, electrical, hardware, software, communication, AI/ML , web skills. So, IoT does not mean any specific technical skill. It is a broad business segment.

Machine learning is quite distinct that way. It signifies a set of technical options, specialized architecture, mathematical algorithms. These are transferable skills from project to project.

From this perspective, there are JDs with IoT "experience". But there is no "one IoT skill". But machine learning JDs are quite clearer because ML is distinct skill. I might hire for simple SVM applications, vision application with deep learning, NLP with DL etc.


Hence my strong recomemndation is :

1. If you want to stay mechanical engineer, search for mech jobs where there are "IoT project". Even otherwise, read up what is IoT. My post is here

2. If you want to be more aligned to software, go for machine learning. You can also use your mechanical experience if there is relevant project (ex: project goal is to reduce mechanical break downs by analyzing data of manufacturing defects or customer recalls).

a. Just do this basic course, if you are not sure. This gives good introduction. Do not be intimidated by mathematics. It is initial necessary evil.


b. If you are almost certain you want to ML, then do this course.


There is also another certification I suggest The difference is the style of presentation. deep starts with mathematics and then gets into practical. The second one starts with practical and then theories as needed.

If you are confused I suggest courses.

c. You also could do python programming course formally. ML guys get rejected for lack of python skills also.




Translate »