Why do so many programs use rational databases instead of loading the data during startup and keeping it in memory? Especially for smaller datasets I would think, that a database adds unnecessary complexity and overhead. Also, a lot of data can be saved using modern RAM and when using an in-memory approach, optimized data structures can be utilised to further improve the performance
I think you answered your own question in a way. For smaller datasets and workloads which only require one machine, storing your data in memory or on disk locally is probably fine.
When you start getting larger datasets or need to have a dataset that’s consistent across multiple servers, you need something more than just the ram of the server.
At that point it’s best not to reinvent the wheel and just use some kind of pre existing datastore who’s tradeoffs make sense for your use case. Some of these datastores do use RAM under the hood for performance but there are pros and cons as with anything.