Schema-on-read

Schema-on-read is a data management approach where the structure or organization of data is defined only when you access or analyze it, not before. Think of it like a library that categorizes books only when you search for a specific topic, rather than organizing all books by subject upfront. This method allows for flexibility in handling diverse, unstructured, or evolving data sources, making it easier to store raw data initially and interpret it later based on your needs. It's commonly used in data lakes and big data environments to adapt quickly to new types of information.