Memory persistence and storage in Agentic AI

Introduction

Memory persistence and storage help AI agents remember important information over time, so they can make better decisions and learn from past experiences.

When you want an AI agent to recall previous conversations or actions.

When storing user preferences to personalize future interactions.

When saving learned knowledge to improve performance over time.

When you need to keep track of long-term goals or plans.

When recovering from interruptions without losing progress.

Syntax

Agentic AI

memory = PersistentMemory(storage_path='memory.json')

# Save data
memory.save(key='user_info', value={'name': 'Alice', 'age': 30})

# Load data
user_info = memory.load(key='user_info')

PersistentMemory is a simple example class that saves data to a file or database.

Use unique keys to store and retrieve different pieces of information.

Examples

Save the last visited page in the session data.

Agentic AI

memory.save(key='session_data', value={'last_page': 'home'})

Load and print user information previously saved.

Agentic AI

user_data = memory.load(key='user_info')
print(user_data)

Remove stored session data when no longer needed.

Agentic AI

memory.delete(key='session_data')

Sample Model

This program creates a simple persistent memory using a JSON file. It saves user info, loads and prints it, deletes it, then shows that the data is gone.

Agentic AI

class PersistentMemory:
    def __init__(self, storage_path):
        import json
        self.storage_path = storage_path
        try:
            with open(self.storage_path, 'r') as f:
                self.data = json.load(f)
        except FileNotFoundError:
            self.data = {}

    def save(self, key, value):
        self.data[key] = value
        with open(self.storage_path, 'w') as f:
            import json
            json.dump(self.data, f)

    def load(self, key):
        return self.data.get(key, None)

    def delete(self, key):
        if key in self.data:
            del self.data[key]
            with open(self.storage_path, 'w') as f:
                import json
                json.dump(self.data, f)

# Create memory instance
memory = PersistentMemory('memory.json')

# Save user info
memory.save('user_info', {'name': 'Alice', 'age': 30})

# Load and print user info
user = memory.load('user_info')
print('Loaded user info:', user)

# Delete user info
memory.delete('user_info')

# Try to load deleted info
deleted_user = memory.load('user_info')
print('After deletion:', deleted_user)

Important Notes

Persistent storage means data stays saved even if the program stops or the computer restarts.

Use simple file formats like JSON for easy reading and writing of memory data.

Always handle cases where requested data might not exist to avoid errors.

Summary

Memory persistence lets AI agents keep important info over time.

Use keys to save, load, and delete data in persistent storage.

Simple file formats like JSON make storing memory easy and reliable.