Memory

Memory integration guide

Memory Management

AIVAH’s intelligent memory system enables your agents to remember key information about users across all platforms and interactions, creating more personalized and contextual conversations.

🧠 How Memory Works

Your AIVAH agent automatically stores and recalls important information during conversations, creating a seamless experience where the agent remembers:

• Personal Details: Names, preferences, and background information
• Conversation History: Previous topics and discussions
• Context Awareness: Relevant details for ongoing projects or relationships
• Cross-Platform Continuity: Information shared across different communication channels

💬 Memory in Action

When users interact with your agent, they can see the memory system working in real-time:

Example Conversation:

• User: “Hey there! I have a couple really happy to chat with you today! What can I help you with?”
• Agent: “Hey my name is James”
• User: “Nice to meet you James! What’s on your mind today?”

The agent automatically stores “James” as the user’s name and will remember this in future conversations.

🎛️ Manage Memory Page

Side nav → Integrations → Manage Memory (or /integrations/manage-memory) lists every long-term memory the agent has stored for each customer.

The table has four columns:

Column	Description
info_key	What was remembered (for example favorite_colour).
info_value	The remembered value.
created_datetime	First time the value was learned.
updated_datetime	Last time the value changed.
Actions	Inline edit and delete on each row.

Export to CSV

Click Download CSV to export the entire table. The file is named customer-information-YYYY-MM-DD.csv, making it easy to drop into CRMs, spreadsheets, or audit pipelines.

📊 Managing Your Memories

The memory management system provides:

Memory Categories

• userName: Personal identifiers and names
• Additional Categories: Preferences, history, context, and more (each row is keyed by info_key)

Memory Actions

• Edit Memory: Modify the info_value inline with the edit icon
• Delete Memory: Remove a row with the trash icon
• Add New: Use the manual entry option to capture important information that the agent did not learn on its own
• Export: Hit Download CSV for a full backup

Memory Details

Each memory entry includes:

• Key (info_key): Type of information stored
• Value (info_value): The actual information remembered
• Timestamps: When the memory was first stored and last updated
• Actions: Edit and delete options

⚙️ Memory Configuration

Automatic Memory Storage

• Smart Detection: Automatically identifies important information
• Context Awareness: Understands when information should be remembered
• Cross-Reference: Links related memories for better context

Manual Memory Management

• Add Information: Manually store important details
• Update Existing: Modify stored information as it changes
• Remove Outdated: Delete irrelevant or outdated memories

🔒 Privacy and Security

Data Protection

• Secure Storage: All memories are encrypted and securely stored
• User Control: Full transparency and control over stored information
• Retention Policies: Configurable retention periods for different memory types

Access Management

• Permission Controls: Define who can access memory information
• Audit Trails: Track memory access and modifications
• Export Options: Download your memory data for backup

🚀 Best Practices

Effective Memory Usage

• Regular Review: Periodically review and clean up stored memories
• Accurate Information: Ensure stored information is current and correct
• Strategic Storage: Focus on information that enhances user experience

Memory Optimization

• Relevant Details: Store information that improves conversation quality
• Context Preservation: Maintain conversation context across sessions
• User Preferences: Remember user preferences and communication styles

📈 Memory Analytics

Track memory usage and effectiveness:

• Memory Growth: Monitor how your agent’s memory evolves
• Usage Patterns: Understand which memories are most frequently accessed
• Conversation Impact: Measure how memory improves conversation quality

🔧 Integration with Other Features

Cross-Platform Memory

• Telephony: Remember caller information across phone conversations
• Chat: Maintain context in text-based interactions
• MCP Apps: Store information from connected applications

Workflow Integration

• Automated Storage: Trigger memory storage through workflows
• Conditional Logic: Use memory in decision-making processes
• Data Sync: Synchronize memory with external systems

🆘 Troubleshooting

Common Issues:

• Memory Not Storing: Check permissions and system status
• Incorrect Information: Use edit function to correct stored data
• Missing Context: Verify cross-platform memory settings
• Performance Impact: Regular cleanup maintains optimal performance

The memory system transforms your AI agent from a stateless responder to an intelligent assistant that builds relationships through remembered interactions.