What it is
Atlas is the reasoning engine that powers Agentforce agents. Unlike simple chatbots that pattern-match responses, Atlas plans and executes multi-step workflows by breaking complex requests into sub-tasks, evaluating which actions to take, and chaining them together. It reasons over your CRM data, knowledge articles, and business rules to make decisions.
Why it matters
Atlas is what separates Agentforce from every other AI assistant. It does not just generate text — it reasons, plans, and acts. This means your agents can handle complex scenarios like qualifying a lead by checking multiple criteria, looking up account history, and scheduling a meeting — all in one interaction.
Key components
- ReAct planning loop
- Tool selection and invocation
- Chain-of-thought reasoning
- Guardrail enforcement
How it connects
Atlas powers every Agentforce agent. When you build topics and actions in Agent Builder, Atlas is the engine that decides which actions to invoke and in what order.
Good to know
Atlas logs its chain-of-thought reasoning, so you can see exactly why an agent made a specific decision — critical for governance and debugging.
Related terms
Agent Governance
The policies, controls, and monitoring systems that ensure AI agents operate safely, compliantly, and within business-approved boundaries.
Agentforce
Salesforce's AI agent platform that enables businesses to build, customize, and deploy autonomous AI agents across sales, service, marketing, and commerce.
Trust Layer
Salesforce's AI trust and safety framework that ensures agents operate within compliance boundaries, prevent misuse, and maintain data security.