withDataStorage

Configures the data storage implementation for persistent state and interaction history.

The data storage component handles all persistence operations including:

• Container state: Serialized AI model weights, parameters, and configuration

• Interaction history: Timestamped user interactions for replay and analysis

• User preferences: Instructions, settings, and customization data

• Performance metrics: Learning statistics and model evaluation data

• Data migration: Schema updates and backward compatibility handling

Storage implementations must guarantee:

• Data isolation: Complete separation between different user accounts

• Thread safety: Concurrent access from multiple coroutines

• Atomic operations: Consistent state even during application crashes

• Encryption: Sensitive data protection using appropriate crypto standards

• Efficient queries: Fast retrieval for real-time prediction scenarios

Common implementations:

• RoomDataStorage: Android Room database with SQLite backend

• SQLDelightStorage: Cross-platform SQL database abstraction

• InMemoryStorage: Volatile storage for testing and development

• EncryptedFileStorage: File-based storage with automatic encryption

Performance considerations:

• Choose storage based on expected data volume and query patterns

• Configure appropriate indexing for frequently accessed data

• Implement data cleanup policies to prevent unbounded growth

• Consider backup and restore capabilities for user data migration

Security considerations:

• All user data must be encrypted at rest using industry standards

• Implement proper key management and rotation policies

• Ensure compliance with data protection regulations (GDPR, CCPA)

• Provide user controls for data export and deletion

Parameters

See also

Throws