Service agents
Resolve common cases and deflect routine questions, 24/7.
Explainer
What is Agentforce?
Insight
Agentforce is Salesforce's platform for AI agents that answer questions and carry out approved tasks using business data. ForceFolks helps teams define agent actions, data access, testing and human handoffs before deployment. Start with one measurable workflow. An agent that can draft a reply should not automatically receive permission to change a customer record or approve a transaction.
Key takeaways
- Agentforce supports AI agents that can answer questions and perform configured actions.
- Data access needs planning: enabling Data 360 and implementing a full data model are different scopes.
- Assistance and action are different: decide which tasks require human approval.
- Pricing depends on the product and usage: review consumption and per-user options against your order form.
What is Agentforce?
Agentforce is Salesforce's platform for building and deploying autonomous AI agents: software that can reason over a request, decide what to do, retrieve data, and take actions (answer a customer, update a record, book a meeting) with guardrails. It is the evolution of Einstein Copilot, which Salesforce renamed Agentforce. See the ForceFolks Agentforce product guide.
How does Agentforce work?
Agentforce runs on the Atlas Reasoning Engine: it breaks a request into steps, retrieves relevant, trusted data (via Data Cloud and RAG), and executes defined topics and actions within Salesforce's trust layer. You configure what an agent can know and do; it supports a choice of large language models (OpenAI, Anthropic, Google).
Before selecting an AI interface, compare Salesforce MCP server types, status and access controls. Local developer tooling is distinct from production access.
What is the difference between Agentforce and Einstein?
Einstein covers predictive and generative assistance, such as scoring and reply drafting. Agentforce supports agents that can perform configured tasks. ForceFolks separates suggested outputs from permitted actions when planning delivery. Check the exact licensed features and approval rules rather than assuming every AI feature can act independently. See Einstein and Agentforce explained.
Use cases
What Agentforce is used for
Sales agents (SDR)
Qualify and nurture leads, book meetings, and prep reps.
Custom agents
Built and grounded on your data and processes.
FAQ
Frequently asked questions
Does Agentforce use ChatGPT?
Agentforce runs on Salesforce's Atlas Reasoning Engine and supports a choice of large language models, including OpenAI (which powers ChatGPT), Anthropic, and Google Gemini, within Salesforce's trust layer. See AI & LLM integration.
How is Agentforce priced?
Agentforce has consumption and per-user options; the applicable charge depends on the product, license and usage type. A conversation is not the same billing unit as an action. ForceFolks separates software consumption from delivery work in its Agentforce pricing guide. Confirm current entitlements and rates against Salesforce's Agentforce pricing and your order form.
Does Agentforce need Data Cloud to work?
Distinguish enabling Data Cloud, now called Data 360, from a full data-unification project. Salesforce's Agentforce data setup guidance describes both an enabled Data Library path and a more extensive implementation. ForceFolks scopes ingestion, identity resolution and data access around the agent's actual task; do not assume every pilot needs the same data project.
Is Agentforce the same as Einstein Copilot?
Agentforce is the rebranded and expanded successor to Einstein Copilot. The agentic capabilities continued and grew under the Agentforce name.
Keep exploring
Keep learning
Make Salesforce work across the business.
Tell us what Salesforce must do. ForceFolks will assess your Clouds, integrations, data, automation, team capacity, and delivery risks. You will receive a recommended path to a working implementation.