In machine learning, features are input in your system with individual independent variables. While making predictions, models use these features. In machine learning, new features can be easily obtained from old features.
Features in machine learning are extremely important as they build blocks of datasets. If the features in your dataset are of quality, the new information you will get using this dataset for machine learning will be of quality as well. Different business problems in different industries should not use the same features.
The most important part of machine learning is feature engineering as it makes a difference between good and bad models.
To perform feature engineering in machine learning you need data scientists or machine learning engineers, who are the data experts and will perform feature engineering right.