/**

Redis is what I use when milliseconds matter — caching, job queues, atomic billing operations, and session management across most of my production systems.

The hot Redis / cold Postgres billing pattern I built for Typcraft is one of the most satisfying architectural decisions I've made. Credit operations execute at sub-millisecond latency through Redis, using Lua scripts for atomic multi-key transactions, while a background outbox processor reconciles with PostgreSQL periodically. Sub-millisecond billing with eventual consistency to the database — it just works.

For Hurivuri, Redis backs all the BullMQ job queues — email delivery, Meilisearch index sync, payment webhook processing, and scheduled maintenance. It also handles session storage for authenticated users.

Moonly uses Redis for NFT collection metadata caching, floor price caching with TTL-based invalidation, and session management for wallet-linked accounts.

Most developers treat Redis as just a cache, but it's a full data structure server — sorted sets, streams, pub/sub, Lua scripting. Using those features properly unlocks real architectural flexibility. Redis is the performance layer that makes the difference between an application that feels fast and one that feels sluggish.

*/