These are incredibly disparate skill sets. Of course anyone would want to hire someone like this, and far more would claim to possess such a broad skill set, but in practice it is extremely rare.

You'd need someone with excellent communication skills (presentation, memo writing, teamwork), project management skills (identifying & overcoming workflow bottlenecks), professional skills (timely responses, political savvy), technical skills (application programming, advanced databases, advanced machine learning, Excel modeling) and finally some business domain knowledge.

This is an uncommon intersection of skills.

> You'd need someone with excellent communication skills (presentation, memo writing, teamwork), project management skills (identifying & overcoming workflow bottlenecks), professional skills (timely responses, political savvy), technical skills (application programming, advanced databases, advanced machine learning, Excel modeling) and finally some business domain knowledge.

This is pretty much the bare minimum requirement for any data scientist job I've ever interviewed for or held.

In my experience, software engineers do not make good business analysts (and data/machine learning engineering is a subset of software engineering). Most business analysts cannot program.

However, it's likely that our experiences simply diverge here.

I'm talking specifically about data science, not business analyst or software engineer.

As the saying goes:

If you're looking for a data scientist with XYZABC skills, that's not a data scientist, that's a data science team.

I've seen this in leadership who want to move to "Devops", its the classic "if we find this one person who can do everything we will have no problems!"

The reality is of course, nobody can be amazing at the full lifecycle of an application. Some do better in infra, some better in backend, front end, etc.

A successful leader must find what is needed for the product/application pipeline and hire appropriate skill sets, trying to find the one candidate to rule them all is giving up on planning IMO.

I, interestingly enough, have that skill set (mostly) and probably a broader set of technical skills than you're imagining. I use it to hire a team of specialists under me and interact with other specialized teams (ie: I speak their language) since I lack depth in too many areas. I wouldn't ever imagine hiring a clone of myself except in cases where I can't build out a larger team for a long period of time.

exactly, these are characteristics of a unicorn and I think most of these skills are trivial to build up over time through practice and self-learning and these skills can yield great benefits both for employers and employees

I guess in a vacuum each of those skills is easy to build up through practice and self-learning (which, lets remember, many people struggle with to begin with). However, I think the fact that you refer to people possessing all of them as "unicorns" should be telling as far as how trivial it actually is to build all these skills beyond a simply passable level.

Or you can recognize that they're the characteristics of a unicorn and split the role into multiple positions.

I think the point is that these skills are not trivial to build up over time

Maybe my sense of terminology is warped, but I always thought of

    DataEngineer = DataJanitor
      ∪ MlEngineer
      ∪ MlOps
      ∪ BusinessAnalyst

Data Scientist is more like some combination of statistician, "whatever ML is if it isn't statistics", lighweight mathematician, data janitor (yes there is overlap), business domain specialist, and code monkey.

Ugh, all this gets you is being mediocre at all of this.

yes, if you need to roll-up everything on your own from scratch.

NO, if you use right amount of automation and software (usable data science workbench with MLOps built in, usable and scalable ETL/ELT framework, usable AutoML, etc, etc.)

You still get someone mediocre at everything just they cover up the gaps for a bit longer. Eventually things they don't understand will interact in ways they don't understand and cause production issues. It's okay to be a generalist, one should however understand the blind spots a generalist has.