Data
Redis
For caches, queues, and the things Postgres shouldn't do.
What it is
Redis is an in-memory data store. We use it for caches, session stores, rate limiting, job queues, and any hot-path that would be wasteful to serve from Postgres.
Why we reach for it
- Microsecond-level reads and writes for hot paths - rate limits, session caches, feature flags.
- Simple, well-understood data structures (strings, lists, hashes, sorted sets) that map cleanly onto real problems.
- Good primitives for distributed locking and coordination.
- Operationally simple - Redis is a single binary and its failure modes are well-known.
When we don’t
- -As the primary store of truth - Redis is memory-first, and durability guarantees are weaker than Postgres.
- -For complex queries - you are not writing joins in Redis.
Where it shows up
Engagements where Redis typically lands.
Also in Data
Working with Redis?
Tell us what you’re shipping.
30 minutes, no pitch deck. We’ll ask what you’re building, what hurts, and whether we’re the right fit. Redis expertise is table stakes - we talk scope, not résumés.