This talk on deploying data science models is technically complex and makes me feel like even basic data science is 5-8 years from maturity.
-
-
Speaker even just said we're 10 years behind modern software dev practices. I agree.
- 1 more reply
New conversation -
-
-
In many ways, it often feels like "we must have this or else we fall behind", with the tools informing the task instead of other way around.
-
With no real way of measuring whether it actually accomplishes anything, or adds any actual value. We simply must do ML *because*.
-
Oh, and often no verification of whether the input, and the associated assumptions, are actually valid, useful, comparable data

-
And if your 'team' is just one or two folks, who's going to explain new, unfavourable insights when CEO walks in three hours before meeting?
End of conversation
New conversation -
-
-
Could be interesting to see if someone could come up with an easily orchestrated prod setup for local/GCP/AWS ML without too much knowledge
-
At the very least, getting that number down to at most two, preferably one (ops person + data scientist, or just ops) for prod deployments.
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.