Memory & State
Clarify one of the most confused areas of agent architecture.
Why This Matters
Teams collapse prompt history, vector memory, and workflow checkpoints into one word. Those are three different stores with three different failure modes.
Architecture Position
Memory sits in the Knowledge Layer; durable workflow state belongs to the Runtime Layer. Mixing them is how agents leak data and lose recoverability. See it on the architecture map.
Start Here
Core Concepts
Concepts
- Context vs memory
- Memory vs state
Memory Types
- Working memory
- Episodic memory
- Semantic memory
- Procedural memory
- Long-term memory
Memory Systems
- Extraction
- Storage
- Retrieval
- Consolidation
- Expiration
State
- Conversation state
- Workflow state
- Agent state
- Checkpoint state
Multi-Agent
- Shared memory
- Isolated memory
- Shared workflow state
Security
- Memory privacy
- Memory poisoning
- Cross-user leakage
Advanced
Advanced material
Deeper guides for this domain are on the way.
Coming SoonArchitecture Patterns
Applied patterns for this domain live in the pattern library.
Browse architecture patternsCase Studies
Enterprise walkthroughs that apply this topic in a full system.
Browse case studiesRelated Topics
Context Engineering
How information is assembled, prioritized, and supplied to models and agents.
Open →Agents & Orchestration
What agents are, where they are appropriate, and how autonomous systems should be structured.
Open →Agent Runtimes & Harnesses
The runtime as the boundary between probabilistic reasoning and deterministic enterprise execution.
Open →AI Security & Trust
Threats created by probabilistic systems that retrieve data and take actions.
Open →Roadmaps
This topic appears in the Enterprise AI Architect path.
Follow the Enterprise AI Architect roadmap