This is a really nice write up/story of an approach to solve the top much data on Postgres problem.

I just came across an architect in a very similar space and workload as instacart busy spending countless hours trying to move off Postgres because they’re at 700GB and it’s not “scaling”.