I get this question almost every day from freshers and students who are wanna-be data scientists or ML Engineers.
So, this is why most data science teams prefer experienced candidates:
Data Science is an art where knowing the theory only takes you so far. One needs to know how models behave in real life and on real data, and how to scale the models while deploying.
These days, thanks to the internet, anyone can write an RNN within minutes. But, only few can explain why they chose an RNN for the problem, instead of, let’s say an LSTM. And out of those, only a handful can successfully deploy them in production.
And those handful are the ones data science teams are after. So, if you’re a student or a fresher, you need to show on your resume that you have worked on a real problem and successfully solved it. Good projects make all the difference.
Don’t do projects like Sentiment analysis on twitter data. You are not solving any real-world problem with that. + I can do sentiment analysis in 2 lines of Python code. So, it’s not a project which would impress a data science team lead.
Something like, Predicting stock trends from sentiment on twitter is a good project, cause it solves a real world problem in investing.
So, if you are a fresher and want to get into a good data science team, have stellar projects on the resume. There’s no other short-cut.