Agent Optimization provides tools to dive deeper into unresolved interactions, identify
knowledge gaps, and analyze agent sessions using the Session Tracing Data Model. Agent
Optimization is a key component of Agentforce Observability, designed to help you understand how
your AI agents perform in real-world scenarios. By analyzing user interactions and agent
responses, you can identify areas for improvement and take action to enhance agent
effectiveness.
Required Editions
| Required Editions |
| Agent Optimization is available in Salesforce Enterprise, Performance, and
Unlimited Editions with Salesforce Foundations or Agentforce 1 Edition. |
Key Features
Agent Optimization introduces several powerful capabilities:
- Intents: Represents a set of interactions within a session that address a specific user
intent or request. Intents are generated daily, then clustered and tagged weekly across
all active agents.
- Quality Scores: Calculated using LLM, quality scores highlight how relevant the agent's
response was to the user's request.
- Session Analysis: Dive into agent-user interactions at the intent level. Understand what
users are asking and how agents are responding.
- Trend Identification: Identify low-performing subagents by quality score, agent
misinterpretations, and areas where conversations weren't handled properly, pointing to
potential configuration gaps.
