This is only a proof of concept.
This PostgreSQL extension provides an aggregate function to compute a Bloom filter as well as a function to query the bloom filter to test whether a value is (probably) in the set.
Example:
SELECT bloom_agg(somecol, m) FROM sometable ...;
(where m is the size of the Bloom filter in bits), write down the result and
DELETE FROM othertable WHERE NOT (somecol << $result_from_above);
The idea is that if othertable is some kind of normalized lookup table referenced by sometable, but it's in a different database (because of sharding or partitioning), then it's hard to clean up unused records efficiently. Using a bloom filter makes this slightly easier. (Of course, this might have concurrency issues, so don't take it as is.)