After attending the Agentforce Now tour in Atlanta, I walked away with a deeper understanding of the transformative potential of Salesforce’s Agentforce, Einstein Bots, and Data Cloud. These innovations are reshaping how businesses interact with both customers and internal teams, blending advanced AI, conversational tools, and robust data management in a way that enhances productivity, personalization, and automation. In this post, I’ll share my key takeaways from the tour and highlight how these tools can revolutionize the way businesses automate workflows, engage with customers, and leverage data to drive smarter decisions.
Agentforce: Merging Salesforce’s core processes with Intelligent Agents
Agentforce is the integration of Salesforce’s key functions and the latest conversational AI features to create ‘smart agents’ that can engage with people in real time. This integration is based on declarative tools such as a prompt builder and agent instructions, on top of business processes written via flows or Apex.
For the customers, Agentforce is an improved kind of chatbot. These smart agents can perform certain tasks, thus being able to converse and help the client easily.
For internal users, Agentforce works as a highly sophisticated Einstein Copilot Agent. This tool can be enhanced with agent instructions and actions to conduct intricate queries and logic that are not easily possible with basic automation tools.
For instance, an agent can suggest appointment schedules for product installation to the customer and also fit them for the product according to their availability, and also ensure that the right technician is assigned and the product is delivered to the client before scheduling the appointment (logic can be written via flow or apex to check agent’s expertise and product’s delivery status ).
Agentforce comes in two types:
Internal (for Salesforce users, not customer-facing)
External (Customer facing, mostly as a chatbot)
Although the tech stack is the same for both internal and external Agentforce, the last one is paid for usage ($2 per conversation). For example, an SDR agent—preconfigured with the right actions to qualify leads—uses LLM to analyze the customer’s messages and respond either autonomously or forward the message to a sales representative. Though external Agentforce is accompanied by certain usage costs, the internal variant of the tool can still duplicate the same functionality with a slightly more manual configuration.
Einstein Bots: The transformation from chatbots to Agents for better interaction with the customers.
For customers, it’s more like the previous version of the product was called Einstein chatbots and now the product has been renamed to Agents. These agents are built on Topics (the needs of the end-user) and Actions (the responses that the user expects). The agent is built on the Atlas reasoning engine, it analyzes the input of the user, determines the Topic, and performs the relevant Actions, which could be an Apex trigger, a flow, or a prompt. Einstein Bots remain relevant in this ecosystem as the most affordable way to engage customers. Although they are still in use, Agents have gone beyond this role to include unstructured conversations, data reading, and writing to the CRM or other systems. They coexist with bots and provide more flexibility on what they can and cannot handle and what they can connect to.
Data Cloud: The system that supports the development of powerful agents.
Data Cloud is an important part of these innovations that helps to combine data easily. Although robust agents can be developed from the primary Salesforce data, Data Cloud enables you to import external data including unstructured data like invoices or contracts in PDF format to support agent actions. It is necessary to enable Data Cloud to enable Agentforce; however, Data Cloud itself is not required and is not charged unless it is used. Data Cloud is an Agentforce enabler by providing a scalable data store and acting as a zero-copy interface to your data lakehouse. In addition, if you enable agent logging, the usage data is saved to Data Cloud and you can track details of agent activity, including the interactions of users, escalations, and others.
The combination of Agentforce, Einstein Bots, and Data Cloud is revolutionizing how businesses automate processes, engage customers, and leverage data. By merging traditional Salesforce features with AI-driven conversational tools, companies can now offer smarter, more interactive experiences across both internal and customer-facing environments. While it’s still evolving, the future of intelligent agents in Salesforce looks brighter than ever!
Happy Agentforcing!