Agentic Memory
Agentic Memory provides persistent recall of past interactions and accumulated knowledge about each individual your agents serve. Data 360 processes agent conversations to extract and store structured knowledge that includes session summaries, profile facts, and preferences. This information builds over time into a comprehensive understanding of each user. Every future interaction builds on previous interactions, regardless of which agent the user interacts with.
REQUIRED EDITIONS
Available in: All editions supported by Data 360. See Data 360 edition availability.
How Agentic Memory Works
When an agent converses with a user, conversation data flows through Data 360’s real-time pipeline.
- The agent sends conversation entries (messages, traces) to Data 360 via the conversation APIs.
- Data 360 persists the conversation data and indexes it for retrieval.
- At a periodic cadence during the session, Data 360 processes accumulated messages to extract memories.
- Data 360 stores extracted memories in the AiParticipantInsight DMO and links them to the relevant individual or unified individual.
- Data 360 makes memories available for context retrieval within seconds of extraction.
All memory is scoped to a data space and linked to a unified individual. A unified individual is a single resolved identity that persists across agents, sessions, and channels.
Memory Types
Data 360 extracts and stores two types of memory. Each type serves a different purpose.
Episodic Memory
Episodic memory captures what happened in a specific conversation. Each extraction cycle produces a compressed summary of the interaction, including topics discussed, decisions made, actions taken, outcomes reached, and key turning points.
Episodic memory:
- Belongs to a single conversation via ConversationId.
- Persists indefinitely unless you delete it.
- Runs per conversation and per agent. Each participating agent produces its own summary.
- Answers the question, “What happened in past sessions?”
Semantic Memory
Semantic memory captures what is true about a user. Data 360 builds semantic memory from conversations over time. Semantic memory includes two categories: profile facts and profile preferences.
Profile facts are durable identity facts about the user, such as name, role, company, relationships, key dates, deadlines, technical environment, and business context. Facts persist indefinitely and build a growing factual record. Profile facts answer the question, “What is true about this user?”
Profile preferences capture stated and inferred user preferences across any subject, such as communication style, product choices, scheduling habits, and tool preferences. Each preference includes a confidence score from 0.0 to 1.0. Preferences decay by default and lose relevance over time unless reinforced. Explicit statements receive higher confidence than inferences, which cap at 0.75. Profile preferences answer the question, “What does this user prefer?”
Semantic memory:
- Links to the individual via IndividualId or UserId, not to a specific conversation.
- Serves any agent in the data space that interacts with that individual.
- Grows cumulatively across all conversations with all agents.
Governance and Access Control
Agentic Memory respects Data 360’s standard governance model.
- User-level access control: Object-level security (OLS), field-level security (FLS), and record-level security (RLS) apply to all memory operations including conversation retrieval, context assembly, and memory extraction.
- Agent-level access control: Determines which agents can read which memories. Each agent reads its own data only.
- Identity scoping: Data 360 links all memory to the unified individual. This ensures consistent visibility across agents. The profile DMO selection in a data space affects all agents in that data space.
See Also
j