A data lake serves as a centralized repository that stores vast amounts of structured and unstructured data in its native format. Its main function is to allow organizations to store data at scale and analyze it as needed. This approach enables data scientists and analysts to leverage large datasets for deep analytics and machine learning purposes without the constraints of traditional database systems.