By: Rafael Balconi

During Dreamforce 2024, Marc Benioff and his team presented Agentforce, what we can call “version 2.0 of Einstein Copilot”. In the session, they demonstrated its potential to revolutionize the way CRM can serve companies. What used to be restricted to just a few internal functionalities can now be extended to the entire CRM target audience, from sales teams to the back office to the customer.

In this new format, Agentforce allows users to use natural language to perform actions in all Salesforce clouds, such as Sales Cloud, Service Cloud and Commerce Cloud. By taking advantage of the MuleSoft APIs and all the metadata used to customize the Salesforce environment, Agentforce offers a flexible and powerful tool for companies.

Unlock the power of Agentforce

AI is increasingly closing the relevance gap between brands and consumers. With Agentforce, companies will be able to provide virtual agents capable of carrying out everyday tasks, such as looking up order information, changing delivery methods and making cancellations. It is also possible for them to ask questions about assistance, guarantees, or to suggest new products for the customer. All this using natural language, connecting all the processes already in place in the CRM with the company’s LLMs of choice.

As companies consider integrating AI into their workflows today and in the future, Salesforce’s recent conference revealed the ability to engage customers across multiple channels, including Slack, SMS and WhatsApp, enabling 24/7 multi-channel service in natural language, providing accurate responses.

Demonstrations at Dreamforce illustrated how Agentforce can accelerate sales processes, quickly resolve service cases and enhance marketing strategies. In addition to seamless integration with CRM data and metadata, it also enables interaction with the Data Cloud for Augmented Recovery Generation (RAG), using structured data from the companies’ Data Lake without copying (Zero Copy), or even using unstructured data, with the use of Vector Database.

By operating on top of the Einstein1 platform, which includes a robust trust layer, it allows companies to implement generative artificial intelligence solutions without the risk of exposing their data. The Trust Layer guarantees the use of LLMs with zero retention.

Agentforce simplifies implementation for companies with quick configuration using pre-built templates and low or no code, allowing easy creation of conversational topics, instructions and actions.

The current scenario and learning from AI

Today’s message on AI for B2B and B2C journeys is clear: make sure you get real, tangible value, without delay. Dreamforce provided great conversations about the need to build customer trust, reimagine interactions and elevate the human aspect of AI creativity, productivity and innovation.

As we enter this new era with Agentforce, companies have a unique opportunity to enhance their operations and customer interactions through intelligent automation.
Embracing these innovations will be key to thriving in a competitive landscape, ensuring that both businesses and customers benefit from advances in AI technology.