Agentforce use cases for employees and customers
Key Takeaway
Employee agents serve internal employees and run as the logged-in user. Service agents serve external customers and partners and run as an agent user, EinsteinServiceAgent User, unless credential-based user verification is turned on. When a Service agent runs as the agent user, the admin must limit its data access on purpose.
Agentforce agents are AI assistants that use generative AI to do a job, or help a person do one, by working through requests in natural language. Some agents help internal employees inside Salesforce, Slack, and mobile. Others talk directly with external customers, partners, and sales leads. For an admin, the main difference between these agents is whose permissions the agent uses when it reads data and takes action.
How agents work
Agents are goal-oriented, autonomous AI assistants built on LLMs. An agent decides what to do by using subagents and actions. It also follows instructions, which act as guardrails so the agent stays on task and does not present false information as fact.
You create a subagent first, and then you create its actions and instructions under it. Both actions and instructions must be linked to a subagent. An action cannot work without a subagent, and a subagent is ineffective without actions.
Some agent types assist and collaborate with a Salesforce user in the flow of work. Other types can also act on behalf of a user or customer, based on the use cases and guardrails that an admin specifies.