Skip to main content

​
Overview

Memory enables agents to remember conversations and learn about users across sessions. It maintains chat history, generates summaries, and builds user profiles for personalized interactions.

​
Installation

Memory requires a storage backend to persist data. Choose the storage option that fits your deployment needs.
Example: Setting up Memory with SQLiteFor local development and testing, use SQLite storage:
uv pip install "upsonic[sqlite-storage]"
Other storage options:
  • [redis-storage] - Redis for distributed, high-performance systems
  • [postgres-storage] - PostgreSQL for production, multi-node deployments
  • [mongo-storage] - MongoDB for document-based, scalable systems
  • [mem0-storage] - Mem0 Platform integration
For production deployments, PostgreSQL or Redis are recommended. See Storage Options for detailed configuration.

​
Key Features

  • Conversation History: Persist complete chat history across sessions
  • Session Summaries: Auto-generate condensed conversation summaries
  • User Profiles: Extract and learn user traits from interactions
  • Multiple Storage Backends: SQLite, Redis, PostgreSQL, MongoDB, or in-memory
  • Sync & Async Support: Both synchronous and asynchronous storage operations
  • HITL Checkpointing: Automatic checkpoint saving for Human-in-the-Loop resumption

​
Quick Start

from upsonic import Agent, Task
from upsonic.storage.memory import Memory
from upsonic.storage.sqlite import SqliteStorage

# Create storage backend
storage = SqliteStorage(db_file="memory.db")

# Create memory with desired features
memory = Memory(
    storage=storage,
    session_id="session_001",
    user_id="user_123",
    full_session_memory=True,    # Enable chat history
    summary_memory=True,         # Enable summaries
    user_analysis_memory=True,   # Enable user profiles
    model="openai/gpt-4o"        # Required for summaries & user analysis
)

# Attach memory to agent
agent = Agent("openai/gpt-4o", memory=memory)

# First interaction
result1 = agent.do(Task("My name is Alice and I'm a Python developer"))

# Second interaction - agent remembers context
result2 = agent.do(Task("What's my name and expertise?"))
print(result2)  # Alice, Python developer

​
Memory Types

TypePurposeRequires
Conversation MemoryFull chat history persistencesession_id
Summary MemoryCondensed session summariessession_id, model
User Analysis MemoryUser profile extractionuser_id, model

​
Storage Backends

BackendUse CasePersistence
SqliteStorageLocal development, single-nodeFile-based
RedisStorageDistributed, high-performanceIn-memory + optional persist
PostgresStorageProduction, multi-nodeDatabase
MongoStorageDocument-based, scalableDatabase
InMemoryStorageTesting, ephemeralNone