What is agent memory?
Agent memory is the information an AI agent keeps so it does not start from nothing on each turn. It spans what happened moments ago in the current task and what the agent learned in past sessions.
The idea
A model on its own is stateless. It answers the prompt in front of it and forgets everything the moment the turn ends. Agent memory is the machinery around the model that carries information forward, so the agent can build on what happened earlier instead of asking again.
Two kinds of memory
Memory splits into two. Short-term memory holds the current task: the steps so far, the tool results, what the user just said. Long-term memory holds what persists across sessions: preferences, past decisions, facts the agent was told once and should not forget.
Why it matters
Without memory, an agent repeats questions, loses the thread on multi-step work, and treats a returning user as a stranger. With it, the agent stays coherent within a task and useful across visits. Storing memory is easy. The hard part is choosing what to keep and what to pull back when.
See it on your own traffic
Add the SDK, define one attribute, and see an attribute updating against your own traffic in the first session. The free tier runs on the real engine: no card, no sales call.