This is not a good use case for Postgres. You might want to look into one of the schemaless solutions, like Mongo. If you're using a single table with lots of d

This is not a good use case for Postgres. You might want to look into one of the schemaless solutions, like Mongo. If you're using a single table with lots of duplicate and empty data, Postgres is going to waste space and processing time. You're not going to use features of or gain any benefit from a RDBMS. You might even be able to get away with using Redis or a similar key-data store, for even better performance. What I really think you need to do is reconsider why you're storing data like this. Anonymous columns, duplicate data, and excessive nulls are all big red flags. If you're not storing relational data, don't use a relational DB. It sounds to me like you might be cramming relational data into a junk bin for what are likely ill-conceived reasons.