Salesforce presented a new portfolio of out-of-the-box AI agents during Dreamforce 2026. These agents are designed to perform work in sales, service, commerce, employee experience, and back office.
Salesforce’s announcement comes ahead of Dreamforce 2026, the company’s annual technology event. With the expansion of Agentforce, Salesforce aims to support companies in deploying AI that goes beyond answering simple questions. The new agents are ready to use and designed to actually perform tasks within diverse business functions.
In-depth knowledge of business context #
The new agents leverage integration with Customer 360, giving them access to customer context and existing business processes within Salesforce. This enables the agents to work purposefully from the start, without the need for extensive configuration. The agents can be deployed in various domains such as sales, customer service, commerce, employee experience, and back-office processes.
The agents do not only perform tasks, but are also capable of collaborating across different business units. When work becomes more complex, they can continue working toward the set goal by combining information from various sources. This makes it possible to automate and optimize processes end-to-end.
Pre-built agents #
Companies no longer need to build their agents from scratch; they can start with pre-built agents designed around specific functions that are adaptable to their own operations. These include:
- Casey for customer service via voice, SMS, WhatsApp, and web chat
- Paige for IT and HR requests
- Carter as a shopping assistant with in-chat checkout
- Hunter for outbound sales
- Marshall for supply chain processes
- Piper for inbound lead generation
- Fin for complex customer experience workflows
Most agents are generally available, while Hunter is still in the pilot phase and will not roll out broadly until November 2026. Customers can tailor each agent to their business, with a custom name and within their own business rules, permissions, and security, and combine AI reasoning with deterministic rules thanks to Agent Script, Salesforce’s open-source language.