heady-redis-lattice

What problem does it solve?

High-latency data access across distributed services is mitigated by orchestrating a three-tier cache hierarchy spanning edge, in-memory, and durable stores, delivering low-latency reads and consistent state.

Core Features & Use Cases

• L1 edge cache (Cloudflare KV) with ~1ms latency
• L2 in-memory Redis (Upstash) for session & rate limiting
• L3 durable Postgres cache with pgvector embeddings
• Phi-scaled TTLs and tiered coherence: write-through for hot data, write-back for warm data
• Pub/sub event bus for cross-service coordination
• Embedding cache and predictive cache warming for hot-path data
• Session store backing with httpOnly cookies and phi-decay eviction

Quick Start

Configure the three-tier cache by wiring your Cloudflare KV, Upstash Redis, and Neon Postgres instances, then perform a get or set to observe L1/L2/L3 behavior.