My dynamodb table is storing multiple attributes on each item, one of which being an array with a relatively large amount of entries. Whenever I read an item of my table, I really only need a certain subset of that array, so it's a waste of data throughput to query the entire thing.
Example: A good analogy would be stock price history; each entry in my table represents a certain stock, and the array attribute is the price history of the stock. When I query a stock, I will always know a start and end date (index of my array), which will usually be a very small subset of the entire array, so I'd optimally like to return my item with just that subset filled in the array attribute.
I guess a more standard way of doing this would be to use a relational database instead with a "prices" table, but that wouldn't fit that well with the rest of my model, while also ending up in being a table with an absolute humongous amount of entries.
Really, what I'm trying to do is just reduce the cost of my AWS calls.