Your question is Approach a Project Problem. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
You're in the middle of a data science project when a meaningful problem appears, such as missing data, a model underperforming, or a dependency slipping. You need to respond quickly without losing sight of the business goal or creating confusion across the team.
How do you approach a problem encountered in a project?