Anonymous
6/24/2025, 9:24:02 PM
No.105692983
>>105689562
It sucks being one of the few competent Indians but yes I recognize everything you say. "Data science" as advertized, i.e. ML engineering is completely useless for the vast majority of companies. The companies where it is useful require you to have publications and preferably a phd. The spots are very competitive.
Doing data wrangling and backend engineering on a database is more lucrative than data science currently is. Except you need to learn a whole lot more to be proficient in that. Honestly knowing about machine learning is a nice-to-have at most companies. If you go into data analysis having a solid grasp of dashboarding, databases, data warehouses, making data pipelines and basic statistics is everything you need. I've only once been asked about something involving machine learning in 5 years of working and it was a small proof of concept about synthetic data. Basic statistical knowledge about distributions, how to interpret company data, how to interpret use cases, how statistical testing / means comparisons work will be far more important than machine learning.
It sucks being one of the few competent Indians but yes I recognize everything you say. "Data science" as advertized, i.e. ML engineering is completely useless for the vast majority of companies. The companies where it is useful require you to have publications and preferably a phd. The spots are very competitive.
Doing data wrangling and backend engineering on a database is more lucrative than data science currently is. Except you need to learn a whole lot more to be proficient in that. Honestly knowing about machine learning is a nice-to-have at most companies. If you go into data analysis having a solid grasp of dashboarding, databases, data warehouses, making data pipelines and basic statistics is everything you need. I've only once been asked about something involving machine learning in 5 years of working and it was a small proof of concept about synthetic data. Basic statistical knowledge about distributions, how to interpret company data, how to interpret use cases, how statistical testing / means comparisons work will be far more important than machine learning.