Netflix Data Scientist Inference Interview
Looking for some intel on the Netflix inference data science internet. At a high level I want to understand the topics covered and the degree of depth. My background is in causal inference and I’ve published and been a research assistant for a few studies leveraging causal inference methods on behavioral data. I feel good about my foundation but I’d love to understand how to better prepare. TC: 415k target (~$300k after stock free fall) #datascience
You'll be fine. Refresh your fundamentals, but more importantly be prepared to think. We tend to ask open ended questions, think about the business need in addition to the technical ask.
is netflix DS phd gated or do people with masters get in as well? Also how fixed is the 5yoe requirement
Not PhD gated, there are DS with master's degrees. We recently opened up lower level roles for less experienced people - but if you fall into this bucket it's going to be harder to compete with the waterfall of PhD grads we get.
Do Netflix have big DS presence in Europe?
No.
Does Netflix still pay cash with the old target TC?
Not 100% sure what you are asking, but yes we still pay all cash.
Thanks. I meant if you were hired with say 400k TC, you are still getting 400k cash despite Netflix stock going down like 70%? This might be hard on the company cost wise especially since they pay top $
Do you have one those damn PhDs?
Is this for the intern role? Or for full time?
Why join sinking ship though?
Who's not sinking? Meta??
😂
What's your yoe ?
Im learning about causal inference. Do you have any links or books you'd recommend on how to build causal inference model?
+1
Book: CI for brave and true
Nobody cares about your target TC
I disagree. In a bear market it is informative if you want to join and think the stock has bottomed out or will only experience moderate declines. In a bull it’s helpful to know what % of the TC is just stock appreciation. I wish it was the norm for people to share both the target and actual TC
(1) Design a causal inference model to to guide hiring decisions. (2) Demonstrate how your particular features are good signals that you’ll be in the top 10 performers using the model that you built in (1)
Thank you! So I’ll demonstrate my technical proficiency in the context of a case study vs a sequence of stats theory questions?
@stitch exactly how DL models being evaluated now a days ... just an empirical proof on age old metrics would suffice.