Expensive database aggregations run on every request without caching

Stats calculations across entire database (4+ years of data) execute on every page load, contributing multi-second delays for data that changes infrequently

Implement time-based caching with TTL (e.g., 24 hours for totals/stats). Options: functools.lru_cache (call-count based), cachetools (time-based), or simple JSON file cache. Check cache freshness before recalculating. For uWSGI multi-process deployments, prefer file-based or external cache over in-memory.