It's still early, but it's been stress-tested pretty aggressively with high write loads and contention scenarios. Right now, it's feature-complete as a high-performance in-memory KV store with dynamic scaling. Clustering and multi-node setups are on the roadmap before any production rollout, but for single-node workloads, it’s already showing strong results (115K+ ops/sec on an M2 MacBook Air).

Are you thinking about a specific use case?