Worked on the xupefei/delta repository to enhance the Delta Lake Spark connector by improving support for generated columns with nested types that differ in nullability from their generating expressions. Addressed schema evolution challenges by implementing type comparison using equalsIgnoreNullability, ensuring robust handling of complex data structures. Improved error messaging for type mismatches, which streamlines debugging and accelerates pipeline development. The changes aligned generated column behavior with top-level type handling, resulting in more consistent and reliable schema management. This work leveraged skills in Data Engineering, Spark, and SQL, and was implemented using both Java and Scala to ensure compatibility and maintainability.