Derived Query Methods in Spring Data JPA — Name-Based Query Generation

Derived query methods are a Spring Data JPA feature where I just write a method signature in my repository interface and Spring generates the query by reading the method name. For example, if I write findByLastName with a String parameter, Spring builds a select with a where clause on the last name field. I don't write any SQL or annotations. This works because Spring parses the method name at application startup. It breaks the name into parts — findBy means it's a select, LastName maps to the last name field on my entity, and the parameter becomes a where condition. I can combine conditions with And or Or, like findByEmailAndStatus. There are operator keywords too, like findByAgeGreaterThan, and result shapers like OrderBy, Top, and IgnoreCase. The mistake Learners make is writing a custom Query annotation for simple lookups that a derived method handles perfectly. Another gotcha is delete methods: if I write deleteByStatus I have to add Transactional, or it throws at runtime. Something interviewers like to hear is that if my method name doesn't follow the grammar or references a field that doesn't exist, the application fails at startup with a clear error, which is much better than discovering it in production. So derived methods keep repositories clean and readable, and when the names get long with many conditions, that's the signal to switch to a Query annotation with JPQL.