Capital One Director, Data Engineering reviews

3 out of the 60 DS actually do work that resembles industry standard DS work. These are the folks that are sent to public and recruiting events to make it sound like we all do real DS work. Don't be fooled! Most of the business leaders do not know what DS resources are or how to use them. They think Data Scientists' core job is to write SQL scripts for business analysts and process managers for "controls". Instead of investing in proper tech resources and fixing processes, leaders choose to use Data Science resources to write SQL script based "controls" designed by business stake holders who know nothing about data or SQL. You don't think you will be writing SQL scripts for data pulling when you sign up for a job description that throws out AI Machine Learning and PhD requirements. It's outright outrageous false advertisement. There are frequent breakdowns in processes that impact our customers. The in-house euphemism for this is a "risk event". Data Science resources are used regularly to fix these risk events by adhoc query writing to pull data for process manager who were not able run their processes properly. No wonder more than a dozen data scientists have left the company in the last few months. Turn over is insanely high. The gave every DS a big fat raise recently hoping to stop the bleeding. Did not help... The kind of problems the business is trying to solve do not require data science, machine learning or even basic statistics. Just some reliable data and SQL developers / Data Analysts would be more than sufficient.

You are hired as a data scientist, but in reality, 99% of times you'd be fixing databases using SQL. Management is always trying out new things, which is good, but that means the end users would always have to constantly debug and validate. You can't really do analysis when the data is bad. Ridiculous number of meetings.