The perception of Redis is that it is fast. It seems like that. But in reality it is not faster than every database or other related software. The only advantage Redis has is it keeps the data in memory and thats all of the magic.
What is a database doing so wrong?
When a query reaches a database it reads data from its store. Sure otherwise data cannot be delivered as response to the asking system. To lookup the correct data a database system uses THE SAME mechanisms as Redis is using. These mechanisms are algorithms like Bloom Filter, HyperLogLog and such. The only difference now is that data which is going to be delivered as response must be read from disk. Redis keeps them all in memory.