Skip to main content

Explore Data Cloud and Agentforce

Learning Objectives

After completing this unit, you’ll be able to:

  • Explain why Data Cloud is required for Agentforce.
  • Identify benefits of building agents that use unified customer 360 data from Data Cloud.

Before You Start

Before you start this module, consider completing this recommended content.

Note

All features described in this module consume Data Cloud credits. Learn more in Data Cloud Credit Consumption: Quick Look and Salesforce Digital Wallet: Quick Look.

Explore How Data Cloud and Agentforce Interact

You took the plunge and got Agentforce in your Salesforce org. You’re excited to build AI agents that know your customers and business. Let’s explore how Data Cloud relates to Agentforce so you understand the connection and how to use them together.

There are two layers of Data Cloud for Agentforce: enablement and implementation.

  • Enabling means you provisioned and enabled Data Cloud in your org.
  • Implementing means you enabled Data Cloud in your org, connected data, and mapped it to data models. You also may have harmonized data with identity resolution rulesets and set up other Data Cloud features.

Enabling and implementing Data Cloud impacts Agentforce capabilities.

  • Data Cloud enabled: Data Cloud must be provisioned and enabled for all Agentforce use. Agentforce features such as the Agentforce Data Library and Einstein Trust Layer don’t work without Data Cloud.
  • Data Cloud implemented: When you implement Data Cloud, you gain benefits such as diverse customer 360 data, powerful data cleansing and transformation capabilities, and fully customizable RAG. To build a robust agent workforce and support complex requirements, you need to implement Data Cloud.

In this module, you learn how enabling and implementing Data Cloud impacts your agents. Data Cloud and Agentforce work together to give you effective, trusted agents grounded in your enterprise data. To understand why you need Data Cloud, you explore how Agentforce capabilities are powered by Data Cloud. You learn if you need Data Cloud enabled or implemented for each capability. You also learn how to implement Data Cloud for Agentforce by following a use case.

You begin by exploring how Data Cloud provides a foundation for effective, personalized agents.

Data Cloud Implemented: Agents Built on Customer 360 Data

When Data Cloud is fully implemented, it offers a robust, secure data foundation that gives agents access to unified customer 360, transformed, real-time, and zero-copy data. In the table, explore some key aspects of a Data Cloud foundation and how they impact agents.

In the NTO Example column, follow an example with Northern Trail Outfitters (NTO), a fictional apparel company. NTO built a service agent that handles customer inquiries, including recommending products, refunding orders, and offering promotions.

Data Cloud Foundation

Description

Agentforce Impact

NTO Example

Unified Customer 360

Bring together data from across Salesforce and beyond, run identity resolution rulesets, and build a unified profile of each customer with full historical context across data sources.

Agents built on Data Cloud data know your customers inside and out, from their purchases in Sales Cloud to their engagements in Marketing Cloud.

Instead of generic or disjointed responses, agents can give personalized responses that consider the customer’s actions across the company.

  1. A customer asks NTO’s agent for a product recommendation.
  2. With unified profiles, the agent can see the customer’s recent purchase: shoes from NTO’s new line. The agent also sees that the customer recently engaged with the line’s marketing campaign, specifically by clicking on the jacket and shoes.
  3. Since the customer already purchased the shoes, the agent recommends the jacket.

Data Transforms

Cleanse and transform data. Resolve data quality issues, such as inconsistent naming and formats.

Clean data improves agent accuracy, consistency, and reliability. Without clean data, the agent is more likely to give incorrect or ambiguous responses, or no response at all.

NTO just ingested order records from Commerce Cloud into Data Cloud. NTO hasn’t had time to clean the data yet.

  1. A customer asks NTO’s agent to refund their order.
  2. NTO’s agent looks up the order number in Data Cloud. However, the date on the order record is in DD/MM/YY format. NTO’s policy is that returns must be processed within 30 days, but the formula only works with MM/DD/YY dates.
  3. NTO’s agent is unable to process a refund.

NTO runs a data transform to convert all dates to MM/DD/YY format. Now agents can successfully process refunds!

Real-Time Data

Ingest, unify, analyze, and act on data in real time.

Create agents that know current customer actions and react in seconds.

  1. As a customer browses NTO’s website, they open the agent chat and ask for current promotions.
  2. The agent sees the customer’s interactions in real time, and recognizes that the customer is clicking on backpacks.
  3. Instead of recommending every active promotion, the agent gives the customer the most relevant promotion: buy one get one on select backpacks.

Zero Copy

Connect data stored outside Salesforce using Zero Copy. Zero Copy lets you create bidirectional communications between Data Cloud and external systems, so you can access data freely without duplicating it.

Expand the reach of your agents beyond Salesforce.

NTO stores loyalty program data in Databricks. They use a zero-copy connection to bring the data into Data Cloud and enrich unified profiles.

  1. A customer asks NTO’s agents for current promotions.
  2. The agent looks up the customer’s unified profile, which has been enriched with loyalty program data.
  3. The agent sees that the customer is a platinum member, which entitles them to an exclusive promotion.
  4. NTO’s agent recommends the exclusive promotion.

While there are other options for ingesting data for Agentforce use, Data Cloud is the only one that supports a holistic 360 strategy with unified data.

  • CRM Data via Fileforce: Only ingests unstructured CRM data, doesn’t unify data.
  • Agentforce via Agentforce Data Library (uses Data Cloud storage): Only ingests unstructured data, doesn’t unify data.

Next Up

In this unit, you discovered how agents built on Data Cloud benefit from access to unified, transformed, real-time, and zero-copy data. After connecting sources and transforming data in Data Cloud, you’re ready to set up Retrieval Augmented Generation (RAG).

Resources

Compartilhe seu feedback do Trailhead usando a Ajuda do Salesforce.

Queremos saber sobre sua experiência com o Trailhead. Agora você pode acessar o novo formulário de feedback, a qualquer momento, no site Ajuda do Salesforce.

Saiba mais Continue compartilhando feedback