So, time to drop some knowledge bombs. Most data scientists aren't taught:
- TCP/IP Protocol architectures
- how to deploy a server
- RESTful vs SOAP web services
- Linux command line tools
- the software development life cycle
- modular functions + the concept of writing tests
- TCP/IP Protocol architectures
- how to deploy a server
- RESTful vs SOAP web services
- Linux command line tools
- the software development life cycle
- modular functions + the concept of writing tests
- distributed computing
- why GPU cores are important
- client-side vs server-side scripting
..and that's just a subset. If you meet a data scientist who has familiarity with those concepts, it's because they either have a CS or IT background, or they taught themselves.
- why GPU cores are important
- client-side vs server-side scripting
..and that's just a subset. If you meet a data scientist who has familiarity with those concepts, it's because they either have a CS or IT background, or they taught themselves.
So be thankful if folks are following along! 😀
And be mindful that sometimes more detailed, patient, lower-level explanations are necessary - especially when writing docs.
R is fantastic at this: for example, @hadleywickham's httr vignette.
cran.r-project.org/web/packages/h…
And be mindful that sometimes more detailed, patient, lower-level explanations are necessary - especially when writing docs.
R is fantastic at this: for example, @hadleywickham's httr vignette.
cran.r-project.org/web/packages/h…
Ex: "you can containerize a model!" or "you can deploy a model as a web service!" or "you can create an API endpoint for your resulting data frame!" do not resonate.
Show examples of what that those options like, in code and output, and why a data scientist would be interested.
Show examples of what that those options like, in code and output, and why a data scientist would be interested.
"Your model could then detect anomalies on a sensor, immediately!" = that works.
"A website or mobile app could access the output of your model based on any given user input." = that works.
"The model you trained on 1000 records could be trained on all billion records." = 👍👍
"A website or mobile app could access the output of your model based on any given user input." = that works.
"The model you trained on 1000 records could be trained on all billion records." = 👍👍
Other thing to note...
It's also very difficult for a DS to know what they *don't* know. They're likely used to being the computer-y one in their research group; so a self-assessment isn't a good gauge.
Ask them to explain things like localhost, HTTP responses, authentication..
It's also very difficult for a DS to know what they *don't* know. They're likely used to being the computer-y one in their research group; so a self-assessment isn't a good gauge.
Ask them to explain things like localhost, HTTP responses, authentication..
...and if they can't, be mindful that some on-the-job training and online courses might be necessary to get them up and running.
Data Scientists are quick studies, I promise! But working with Meg to synthesize golden nanoparticles or w/e != deploying a model to production.
Data Scientists are quick studies, I promise! But working with Meg to synthesize golden nanoparticles or w/e != deploying a model to production.
Also excellent, excellent points. Version control, pair programming and code review, Agile project methodologies, what a "remote virtual machine" actually is and why you would want one -- all of those concepts are probably going to need to be explained.
