Memory
Memory allows agents to retain information across conversations, enabling personalized experiences and continuity with returning callers.How Memory Works
Memory Types
Short-Term Memory
Context within a single conversation:Long-Term Memory
Persists across conversations:Workspace Memory
Shared knowledge across all agents:Configuring Memory
Automatic Capture
Automatically extract and store information:Manual Storage
Store information via pathway actions:API Storage
Store memory via API:Retrieving Memory
Automatic Retrieval
Memory is automatically available in conversations:Via API
Memory in Conversations
Personalized Greetings
Context Continuity
Preference Application
Memory Schemas
Define what can be stored:Privacy Controls
Retention Policies
Caller Opt-Out
Allow callers to request data deletion:Data Access
Provide access to stored data:Anonymization
Memory Events
Track memory operations:| Event | Description |
|---|---|
memory.stored | New data stored |
memory.retrieved | Data retrieved for call |
memory.updated | Existing data modified |
memory.deleted | Data deleted |
memory.expired | Data auto-deleted |
Webhook Payload
Use Cases
VIP Recognition
Follow-Up Context
Preference Learning
Best Practices
Only store what's useful
Only store what's useful
Don’t store everything—focus on data that improves the experience.
Respect privacy
Respect privacy
Be transparent about data storage and honor deletion requests.
Set appropriate retention
Set appropriate retention
Don’t keep data longer than necessary.
Handle missing memory gracefully
Handle missing memory gracefully
Always have fallbacks when memory data isn’t available.
Test personalization
Test personalization
Verify personalized experiences work correctly with real data.
Next Steps
- Agent Personas — Create personalized interactions
- Data Handling — Understand data policies
- Webhooks — Track memory events