A friend of mine finished her data science bootcamp last October.
She could build an end-to-end scikit-learn Pipeline in her sleep.
Six months into her first job, she’d used it exactly once.
“I feel like I forgot everything.”
She hadn’t forgotten anything. The stack just changed the moment she stopped interviewing.
The Python libraries you learn for interviews are not the libraries you use at work.
Not because interviews are outdated. Because they’re testing for something different.
Interviews test one thing:
Can you build the ideal version?
Work rewards a different one:
Can you make the messy version work with what’s already there?
The interview stack is clean. scikit-learn Pipelines. PyTorch model classes. SQLAlchemy. dbt. MLflow. Every library used the way its documentation intended.
The work stack is whatever the team had when you joined. pandas doing eighty percent of the load. A utils.py file three engineers ago wrote. One SQL client. A script called run.py that nobody has time to refactor.
The library the team already uses will always beat the one you’re better at.
Three things nobody tells juniors about this gap:
Ninety percent of your real value lives in pandas, SQL, and knowing where to look. The rest is optional most weeks.
Your fancy stack is what you build to, not what you build with. You earn the room to change how the team works before you use it.
Feeling behind because you’re “only using pandas” means you’ve misread what work rewards. You’re not behind. You’re doing the job.
Interview prep isn’t wasted, it buys you the interview.
But once you’re inside, the person who ships in the ugly stack the team already trusts will always be more valuable than the one who knows a prettier one.
Interviews reward the ideal. Work rewards the workable.
Learn both. Use each in the room it belongs in.
— Josep
Are you still here? 🧐
👉🏻 I want this newsletter to be useful, so please let me know your feedback!
Before you go, tap the 💚 and the restack buttons at the bottom of this email to show your support, it really helps and means a lot!
Which library did you spend the most time learning for interviews and use the least at work? Please let me know👇🏻