What you need to know

Since its launch at the Dreamforce conference in September 2024, Salesforce Agentforce has changed the conversation about artificial intelligence, customer experience and customer service. Shortly after the launch, Salesforce announced that it was hiring more than 1,000 employees to meet the demand for Agentforce.

At the time of Agentforce’s announcement, Salesforce CEO Marc Benioff called Agentforce “the third wave of artificial intelligence — advancing beyond co-pilots into a new era of highly accurate, low-hallucination intelligent agents that actively drive customer success.”

The platform is so important to Salesforce that Benioff said when introducing Agentforce: “The only thing we’re going to do at Salesforce is Agentforce.”

What is Agentforce?

Agentforce is essentially a platform layer in the Salesforce ecosystem. Agentforce bots are generative AI bots, but unlike popular genAI platforms like ChatGPT and Google Gemini, Salesforce says its Agentforce bots can be trusted to take action on their own.

Unlike earlier chatbots, AI agents do not follow a prescribed pattern of responses to common queries, but are able to respond more flexibly to questions or problems because they are trained on internal data and are able to use the support of a large language model (LLM). Salesforce’s agent AI, for example, can be trained on data in the Salesforce Data Cloud.

Agentforce combines three major Salesforce tools — Agent Builder, Model Builder, and Prompt Builder — to deliver out-of-the-box bots that work across industries. Bots are customized using low-cost, low-code tools.

Initial agents created using Agentforce include:

• Campaign optimization tool, which manages and implements full life cycles of campaigns.
• Service agentwhich replaces chatbots in handling customer service.
• Bather, which helps B2B customers find products, make purchases and track orders.
• Personal buyerwhich recommends products and helps with searches.
• Traderwhich performs all website-related tasks for merchants, including promotions, product descriptions and insights.
• Sales Development Representativethat engages with potential leads 24/7.
• Sales trainerwhich helps train sales teams and allows salespeople to practice pitching.

General availability for Agentforce One began on October 25, 2024, priced at $2 per call, with volume discounts available

Salesforce also announced the Agentforce Partner Network at the launch. Enables other organizations to deploy third-party agents or use third-party actions when designing custom agents in Agentforce.

Dig deeper: Customer Experience Management in the Age of Agent Artificial Intelligence

How does Agentforce work?

During the first Agentforce announcement at Dreamforce, Benioff said that Agentforce is designed for users who don’t know much about artificial intelligence or building robots.

“They won’t need to be experts in the LLM, they won’t need to understand all these deep skills that you need to know to be a computer scientist,” he said. This is “AI for the rest of us.”

Agentforce is powered by an upgraded version of Salesforce’s Einstein AI called the Atlas Reasoning Engine, which is “designed to simulate how humans think and plan.”

In the announcement, Salesforce made a comparison to self-driving cars. Agentforce can interpret data to adapt to real-time conditions and can operate independently within the company’s hedges.

Benioff said Agentforce has the lowest hallucination rate of any generative AI, but he didn’t say what that rate was.

Improved Atlas Reasoning Engine

Salesforce announced in December 2024 that its Atlas Reasoning Engine has been enhanced with Advanced Reasoning and Augmented Retrieval (RAG), enabling agents to process more complex questions in multiple steps and provide more accurate and precise answers.

Enhanced RAG capabilities allow users to verify information by providing inline quotes with agent responses. Citation of accurate sources builds trust in AI results. It also helps users to verify accuracy.

Data Cloud

Data Cloud provides the user data that underpins Agentforce, making agents more contextually aware, knowledgeable and responsive to user needs. The Data Cloud connects, aggregates and harmonizes customer data and metadata, giving agents access to the accurate information they need to deliver precise, contextual responses. It turns structured and unstructured data – including emails, transcripts and PDFs – into easily digestible Large Language Models (LLM). RAG functionality allows users to upload files and URLs as data sources for their agent.

Customer 360

Agentforce is deeply integrated with Salesforce Customer 360, leveraging applications such as sales, service, marketing and commerce. An end-to-end view of the customer is key to seamless handover to human agents with full conversation history.

MuleSoft

MuleSoft is a platform for API integration, automation and management. It extends the capabilities of Agentforce by enabling Salesforce developers and administrators to leverage its APIs and ingest third-party data.

Slack

Slack is the central point for employee interaction with Agentforce. They can access information, launch workflows, and manage tasks right within Slack, meaning they don’t have to jump from one app to another.

Agentforce also uses Slack Enterprise Search to learn from conversations in Slack channels and direct messages. This helps agents provide more relevant responses and actions based on the context of the conversation.

Tableau

Tableau Einstein is an AI-powered visual analytics platform that includes Agentforce. It’s designed to intelligently incorporate data and insights into workflow and empower employees to use data to drive action.

Dig deeper: 4 Key Features in Salesforce Agentforce 2.0

Agentforce Testing Center

In December 2024, Salesforce announced a Test Center that allows companies to test their AI agents before deployment. The Test Center gives Agentforce users the opportunity to ensure that agents are following instructions, sticking to the facts and working fast. This helps reduce risk and ensures consistent and reliable agent performance. It also provides tools that make it easy to test, audit, and scale agents.

Skills library and integration

Also announced in December 2024, this library includes key business applications such as CRM, Slack, and Tableau, as well as integrations from AppExchange partners. These ready-to-use skills simplify the process of building custom AI agents. The library has pre-built Slack actions like Create Canvas and Message Channel, which allow teams to create agents for their Slack workflows.

Here are some examples:

• CRM skills allow you to create agents to automate and improve user experience. These agents can do things like lead nurturing, sales training, and scheduling service engagements among other functions.
• Sales development skills creates agents that nurture leads based on predefined rules.
• Sales coaching skills enables agents to actively participate in sales calls, providing real-time feedback to reps and improving their performance.
• MuleSoft integrations some companies use Agentforce throughout their business ecosystem. For example, it can connect Agentforce to its ERP system to automate the order fulfillment process. It can also be integrated with marketing automation platforms to launch personalized email campaigns based on customer interactions.
• AppExchange augments the capabilities of Agentforce by allowing users to access specialized skills created by partners.

Which companies use Agentforce?

Salesforce itself uses Agentforce internally. It says that agents on help.salesforce.com now handle 83% of customer support inquiries on their own, with human escalations falling by 50% in the two weeks after implementation.

RBC uses Atlas and Data Cloud with Agentforce to help its financial advisors. Salesforce Data Cloud integrates all RBC client data into a unified system, giving advisors a comprehensive view of their clients. Atlas can answer complex client questions using multilevel inference. It also uses RBC data and business logic to provide a reliable answer.

Accenture uses Agentforce to give employees quick notifications about important accounts. It helps automate document retrieval and provides recommendations for solving business problems. It also streamlines proposal creation and visualization and improves employee collaboration by providing real-time updates, document retrieval, proactive recommendations and proposal creation.

Other companies publicly identified by Salesforce as Agentfore customers include:

• Adecco Group
• BACA systems
• OpenTable
• Sax
• Wiley

Dig deeper: SharkNinja embarks on its Salesforce AI journey