Metadata
- Author: Super Data Science Podcast with Jon Krohn
- Full Title:: 665: How to Be Socially Impactful and Financially Successful in Your Data Career — With Josh Wills
- Category:: 🗞️Articles
- URL:: https://m.youtube.com/watch?v=Hng2RmbSkQE&utm_source=substack&utm_medium=email
- Finished date:: 2024-03-21
Highlights
the infinite Loop ofsadness the infinite Loop of sadness yesyes is is it a part of the infinite Loopof sadnessum yes it is it is it is very much to anextent um I thinkthe the idea of I guess for folks who’venot heard of the infinite Loop ofsadnessumis that the stakeholders that kind ofrevolve around the data of a company um generally report up through differentparts of the org so you might havebusiness users reporting up through somethrough Finance like sort of GNA kind ofstuff some through engineering somethrough product uh some through customersuccess whatever there’s all thesedifferent stakeholders for dataum there’s a data science team somewherethat wants to do sort of cool datascience stuff wants to get to like a btesting experimentation contextualBandits machine learning problemsthere’s a data engineering team whichmay or may not be part of the same organization like could be like at slackfor instance like analytics reported upthrough product at first and then reportit up through GNA but data engineeringalways reported up through engineeringand so they’re worried about performancethey’re worried about cost they’reworried like especially these days likeCloud savings all that kind of stuffand then like you have the rest ofengineering which is like you knowtrying to like terraform things andmonitor things and observe things andkeep the basic product up and runningand stuff like that and so and then likeoh yeah there’s this data warehousething on the side that we also kind of need to like provision some resourcesfor and you know terraform or whateverrightum and yet like yeah like the artifactof having the result of having all thesedifferent stakeholders in thesedifferent organizations with differentobjectives with different goals is kindof this messum that leads from like business peoplehiring data scientists data scientistssay we need data Engineers dataEngineers say we need machines the restof engineering says okay Finance we needa bunch more money to pay for all thesemachines business is confused because they’ve hired all these people and spentall this money and don’t have anyanswers to their questions yet and stuffand so these these things kind of spiralout of controlum I don’t have a great answer to thisunfortunately I’d like to say that I dolike it’s it’s something to identify aproblem I guess where I’ve seen it workwell or the best places I’ve seen itwork is where datais centralized in a single organizationunder a single leader who has like thetrust and confidence of the rest of the executive teamum this doesn’t always happen in fact ithardly ever happens and even when itdoes happen it’s kind of an unstablesituation because uh once that leaderleaves it tends to be the case the wholething kind of falls apart (View Highlight)