You are designing a data model for AWS Lake Formation to store log data from various applications. The log data includes timestamps, log levels, messages, and application IDs. The model must support schema evolution to accommodate changes in log data formats. Which of the following strategies would be most effective?