Data Science Team
topic on 3 shows · 6 statements across 6 episodes
Capital Allocators 
the MAD Podcast 
the a16z Podcast
6 statements about Data Science Team, every show
Rees: In-house data science is inefficient even for top asset managers
“It's not really economical or efficient for even the largest firms that we're partnered with to have their own data science team.”
Hanlon: Reddit uses a hub-and-spoke model for data science
“Data science is hub and spoke. So so effectively that is all of those folks report into the director of data science, who's, who's one of my team members and, but they actually, when we're in a, when we're in an office setting, they sit with the downstream tea…”
Netflix data scientists are free to choose their own tools
“Netflix has this really interesting corporate culture of freedom and responsibility. Which means that our data science teams, they are essentially free to use whatever tooling that works best for them.”
Doshi: Data science teams should only handle 10% of business questions
“90% of a business's questions are pretty reasonably easy to get to. It's just the data and the analysis, and there's some edge cases that you have to kind of figure out. That part's annoying and hard, but 90% of your questions are pretty easy. There's this 10%…”
Bloom: Very few data science teams prioritize model explainability
“Explainability or interpretability turned out to be a very, very important optimization that very few data science teams will be cognizant of unless they're really thinking about it.”