Memory
Agents use memory to store chat history, user preferences and conversation summaries.
Built-in Memory
Every Agent comes with built-in memory that can be used to access the chat history and run messages, but this only lasts while the session is active. Access it using agent.memory
- Chat History: The messages between the User and Agent.
- Run Messages: The full list of messages sent to the model, including system prompt, tool calls etc.
Example
from rich.pretty import pprint
from phi.agent import Agent, AgentMemory
from phi.tools.yfinance import YFinanceTools
agent = Agent(tools=[YFinanceTools()])
# -*- Print a response
agent.print_response("What is the price of AAPL?", stream=True)
# -*- Get the memory
memory: AgentMemory = agent.memory
# -*- Print Chats
print("============ Chats ============")
pprint(memory.chats)
# -*- Print messages
print("============ Messages ============")
pprint(memory.messages)
Persistent Memory
Built-in memory only lasts while the session is active. To persist memory across sessions, we store Agent sessions in a database like PostgreSQL using AgentStorage.
Storage is a necessary component when building user facing AI products as any production application will require users to be able to “continue” their conversation with the Agent.
The general syntax for adding storage to an Agent looks like:
from phi.agent import Agent
from phi.model.openai import OpenAIChat
from phi.storage.agent.postgres import PgAgentStorage
agent = Agent(
model=OpenAIChat(id="gpt-4o"),
storage=PgAgentStorage(table_name="agent_sessions", db_url="postgresql+psycopg://ai:ai@localhost:5532/ai"),
# This will automatically add the chat history to the messages sent to the model
add_history_to_messages=True,
# Number of historical responses to add to the messages.
num_history_responses=3,
)
Read more in the storage section.
User preferences and conversation summaries
Along with storing chat history and run messages, AgentMemory can be extended to automatically classify and store user preferences and conversation summaries.
To do this, add a db to AgentMemory and set create_user_memories=True and create_session_summary=True
User memories are stored in the AgentMemory.db whereas session summaries are stored in the AgentStorage table with the rest of the session information.
Example
from rich.pretty import pprint
from phi.agent import Agent, AgentMemory
from phi.model.openai import OpenAIChat
from phi.memory.db.postgres import PgMemoryDb
from phi.storage.agent.postgres import PgAgentStorage
db_url = "postgresql+psycopg://ai:ai@localhost:5532/ai"
agent = Agent(
model=OpenAIChat(id="gpt-4o"),
# Store the memories and summary in a database
memory=AgentMemory(
db=PgMemoryDb(table_name="agent_memory", db_url=db_url), create_user_memories=True, create_session_summary=True
),
# Store agent sessions in a database
storage=PgAgentStorage(table_name="personalized_agent_sessions", db_url=db_url),
# Show debug logs so you can see the memory being created
# debug_mode=True,
)
# -*- Share personal information
agent.print_response("My name is john billings?", stream=True)
# -*- Print memories
pprint(agent.memory.memories)
# -*- Print summary
pprint(agent.memory.summary)
# -*- Share personal information
agent.print_response("I live in nyc?", stream=True)
# -*- Print memories
pprint(agent.memory.memories)
# -*- Print summary
pprint(agent.memory.summary)
# -*- Share personal information
agent.print_response("I'm going to a concert tomorrow?", stream=True)
# -*- Print memories
pprint(agent.memory.memories)
# -*- Print summary
pprint(agent.memory.summary)
# Ask about the conversation
agent.print_response("What have we been talking about, do you know my name?", stream=True)
Attributes
| Parameter | Type | Default | Description |
|---|---|---|---|
memory | AgentMemory | AgentMemory() | Agent’s memory object used for storing and retrieving information. |
add_history_to_messages | bool | False | If true, adds the chat history to the messages sent to the Model. Also known as add_chat_history_to_messages. |
num_history_responses | int | 3 | Number of historical responses to add to the messages. |
create_user_memories | bool | False | If true, create and store personalized memories for the user. |
update_user_memories_after_run | bool | True | If true, update memories for the user after each run. |
create_session_summary | bool | False | If true, create and store session summaries. |
update_session_summary_after_run | bool | True | If true, update session summaries after each run. |