Vertex AI Agent Engine Sessions maintains the history of interactions between a user and agents. Sessions provide definitive sources for long-term memory and conversation context.
You have several options to use Vertex AI Agent Engine Sessions:
Agent Development Kit: Once you deploy your Agent Development Kit (ADK) agent to Vertex AI Agent Engine, session management is handled automatically.
API calls: You can make direct API calls to Vertex AI Agent Engine Sessions if you don't want to manage sessions using an ADK agent.
Core concepts
Session: A session represents the chronological sequence of messages and actions (events) for a single, ongoing interaction between a user and your agent system.
Event: An event stores the content of the conversation, as well as the actions taken by the agents such as function calls.
State: A state holds temporary data relevant only during the current conversation.
Memory: Memory is personalized information that can be accessed across multiple sessions for a particular user. The agent can use memories to personalize responses to the user and ensure cross-session continuity.
Core functionalities
The core functionalities of Vertex AI Agent Engine Sessions include the following:
Starting new conversations: Create new sessions when a user begins an interaction with an agent.
Resuming existing conversations: Retrieving a specific session so the agent can resume a conversation that has been paused.
Saving progress: Append new interactions (events) to a session's history to update the session.
Listing conversations: Find the active session threads for a particular user and application.
Cleaning up: Delete session objects and their associated data when conversations are finished or no longer needed.