5 ms·

I have a love / hate relationship with bucketing a whole lot of formerly separate disciplines into the data science label. In my case, I'm a reasonably solid R

I have a love / hate relationship with bucketing a whole lot of formerly separate disciplines into the data science label. In my case, I'm a reasonably solid R / Python programmer (who occasionally dabbles in racket, clojure, and others). I've got the sort of applied statistical training of someone who took a quant-heavy course load in a PhD program. I've even (by title) been a data scientist a couple times now. Having recently decided to reenter the job market, I'm reminded that finding the RIGHT data science role is a major challenge. When someone wants a data scientist, they may be looking for someone with a lot of specialist depth in operations research, financial forecasting, machine learning, data-focused software engineering, or some other not-at-all-universal area of expertise. In some ways, I almost think that someone with a bootcamp level understanding of stats may be at an advantage. Whereas I'm very inclined to be, "Oh! You want this other kind of person. Would you like me to put you in touch with one?" I think someone more junior is inclined to be "How hard can [X] be?"

posix_compliant 8y ago
As a "junior" applying to data science positions, I have a 50/50 mix of "no way I'm qualified for that", and "that sounds like a really fun problem."
- peatmoss 8y ago
  Welcome to impostor syndrome. It NEVER goes away. The worst part of impostor syndrome is that sometimes, you really aren't qualified, and so you can never fully move past the self-doubt. You don't sound like the Dunning-Kruger sort, so I'd chase the fun sounding problems in organizations where you can use more senior people that you respect as sounding boards / mentors. Good luck (to us all)!
brational 8y ago
To be fair though, can't you say the same about the title "software engineer". Without further qualifiers such a title can run quite the gamut of skill-sets.
- peatmoss 8y ago
  I see more segmentation in software roles: front-end, back-end, full-stack, embedded, devops, realtime systems are all labels that get you in roughly the right place. Beyond that, there are labels that refer to specific technologies that get you even closer: Rails, Elixir, Django, React, QT, C#/Winforms, Verilog. There really aren’t the equivalent shorthand filters in data science.
  - achompas 8y ago
    > There really aren’t the equivalent shorthand filters in data science. There are, though, right? The catch-all "Data Scientist" title is being split into Data Engineer, Machine Learning Engineer, Data/BI Analyst, etc. Each of those, in my experience, have clear definitions.

posix_compliant 8y ago

As a "junior" applying to data science positions, I have a 50/50 mix of "no way I'm qualified for that", and "that sounds like a really fun problem."

peatmoss 8y ago
Welcome to impostor syndrome. It NEVER goes away. The worst part of impostor syndrome is that sometimes, you really aren't qualified, and so you can never fully move past the self-doubt. You don't sound like the Dunning-Kruger sort, so I'd chase the fun sounding problems in organizations where you can use more senior people that you respect as sounding boards / mentors. Good luck (to us all)!

brational 8y ago

To be fair though, can't you say the same about the title "software engineer". Without further qualifiers such a title can run quite the gamut of skill-sets.

peatmoss 8y ago
I see more segmentation in software roles: front-end, back-end, full-stack, embedded, devops, realtime systems are all labels that get you in roughly the right place. Beyond that, there are labels that refer to specific technologies that get you even closer: Rails, Elixir, Django, React, QT, C#/Winforms, Verilog. There really aren’t the equivalent shorthand filters in data science.
- achompas 8y ago
  > There really aren’t the equivalent shorthand filters in data science. There are, though, right? The catch-all "Data Scientist" title is being split into Data Engineer, Machine Learning Engineer, Data/BI Analyst, etc. Each of those, in my experience, have clear definitions.