Salesforce, a cloud CRM company, said on Wednesday it has unveiled Agentforce Help Agent, designed to reduce the burden of building and operating AI agents for customer service.

The company said Agentforce Help Agent is based on Salesforce's Agentforce 360 Platform and supports service operations by using a company's customer response policies and business information to respond to customer inquiries and help manage each case.

It is designed to handle follow-up tasks linked to customer service beyond simple question-and-answer exchanges. In addition to handling basic inquiries and case management, companies can add functions to process customer requests such as order management, appointment scheduling and account management. It supports delivering a consistent customer experience across communication channels including voice, web, portals and messaging.

Salesforce also highlighted a "pay-per-resolution" billing model.

Companies pay only when an AI agent autonomously resolves a customer issue from start to finish. They do not pay if a customer requests to be connected to a human agent or expresses dissatisfaction with the result. Salesforce also stressed that no separate charges apply for Data 360 and use of Agentforce generated during conversations between the agent and the customer.

Salesforce also said it has unveiled Agentforce Customer Service Portal to make it easier for companies to build self-service environments for their customers. Customers can enter the content of their inquiry on the portal, review a personalised answer and proceed with any necessary follow-up actions.

Kishan Chetan (키샨 체탄), executive vice president and general manager of Agentforce Service at Salesforce, said the core value of AI agents in customer service goes beyond fast answers to resolving customer issues through to the end. He said Agentforce Help Agent draws on experience accumulated over the past 2 years in real operational environments to help companies more easily deliver personalised and proactive omnichannel customer experiences. He said the pay-per-resolution model helps enhance the value of AI-based customer service and improve investment efficiency by using actual service performance as the benchmark.