Skip to main content
A Learning Machine coordinates multiple stores, each optimized for a specific type of knowledge.
StoreWhat it capturesScopeUse case
User ProfileStructured fields (name, role, preferences)Per userPersonalization
User MemoryUnstructured observations and factsPer userContext retention
Session ContextGoals, plans, and progressPer sessionLong-running tasks
Entity MemoryFacts about external entitiesConfigurableKnowledge graphs
Learned KnowledgeInsights that transfer across usersConfigurableTeam-wide improvement
Decision LogDecisions with reasoningPer agentAuditing and learning

Guides

User Profile

Structured facts about users

User Memory

Unstructured observations from conversations

Session Context

Goals, plans, and progress for active sessions

Entity Memory

Facts about companies, projects, people

Learned Knowledge

Insights that benefit all users

Decision Log

Decisions with reasoning for auditing