Performant database tree traversal with Rails

We solved a performance problem in our Rails API when performing a tree traversal by creating an in-memory cache to store the snapshots and preloading it with recent snapshots from each branch. This reduced the number of database queries from thousands to just a few, solving the N+1 query problem. We also tested and refined the solution using feature flags and data on hits and misses, and considered alternative solutions such as using a recursive common table expression or materialized path in the database.