Context architecture is replacing RAG in AI

Redis launched Iris, a context and memory platform for AI agents that automates data access and persists agent memory.

Iris combines continuous ingestion, semantic retrieval, low-cost storage, and runtime access to structured data for enterprise AI.

The move positions Redis’s context architecture as a layer beyond traditional RAG for agent-heavy workloads.

It also reflects a broader enterprise shift toward hybrid retrieval and custom in-house stacks for AI agents.