I think you should start by checking with development if they can split one big feature to many small features with low dependency. If yes, when they submit each small feature to tester, you can test it follows best case as I mentioned.
Regression test manually or automation should be done if they have released some features. It will help you to ensure that it really has no impact between released features.