Agent memory
The persistent store an AI agent uses to retain and recall information across an ongoing task or conversation — distinct from one-off retrieval (RAG).
Agent memory is what lets an AI agent remember — across a long task or many sessions — rather than starting fresh each time. By 2026 it had become a real engineering discipline, with distinct hosting models: managed cloud, self-hosted open source, and local (often via MCP).
It’s related to but not the same as RAG: RAG retrieves from a static knowledge source on demand, while memory continuously incorporates new information from the agent’s ongoing interaction. For audio, a transcribed, embedded archive in a vector database is the substrate that agent memory reads from.