
Having access to Agentforce Knowledge can significantly enhance your service, allowing you to respond quickly and accurately to customer inquiries.
By leveraging Agentforce Knowledge, you can tap into a vast repository of information, streamlining your workflow and improving overall efficiency.
This knowledge base is built on a foundation of structured data, making it easy to search and retrieve relevant information.
With Agentforce Knowledge, you can provide personalized support to customers, addressing their unique needs and concerns.
By doing so, you can build trust and loyalty with your customers, setting your business apart from the competition.
You might enjoy: Journal of Information & Knowledge Management
Agent Configuration
To configure Agentforce to use Knowledge, you'll need to complete a few steps. First, go to Setup > Agents and click "New Agent" to start the setup Wizard.
Make sure to add the General FAQ topic to the agent on screen 2 of the Wizard. Note that if you miss this step, you can add a topic after the agent has been created.
To create an Agentforce agent, follow these five steps: Clarify Roles, Identify Critical Data, Define Instructions & Guardrails, Develop Actions, and Select Channel.
Here's a breakdown of each step:
- Clarify Roles: Define the use case and purpose of the agent, and determine topics, KPIs, and a scope for the agent's actions.
- Identify Critical Data: Determine the data the agent will need to successfully complete desired actions, and check if the data is complete and ready to use.
- Define Instructions & Guardrails: Set operational guidelines for agents by defining core instructions for each topic.
- Develop Actions: Use Apex, Flows, and Prompts to build automations that your agent can call upon to achieve its designated goals.
- Select Channel: Expose your Agent to appropriate channel(s) so they can fulfill their designated purpose.
By following these steps, you can create a Salesforce Agent that can autonomously pivot between distinct tasks to resolve all aspects of a basic customer service engagement without needing to loop in a human representative.
Here's an example of how this works in practice:
Agent Building and Customization
Building an Agentforce agent requires careful configuration and customization. To start, you'll need to configure the agent and topic, which involves adding the General FAQ topic to the agent during the setup process. This is a crucial step, as it enables the agent to access the knowledge base and provide accurate responses.
If you missed this step during the setup process, you can add the topic later by clicking "Open in Builder", adding a topic, and assigning the standard "Answer Questions with Knowledge" action to the topic. This will allow the agent to access the knowledge base and provide responses.
A different take: Personal Knowledge Base
To ensure the agent can access the knowledge base, you'll also need to grant the Agentforce Data Library user the necessary permissions. This involves creating a new Permission Set with object- and field-level access to the Knowledge__kav object and assigning it to the Agentforce Data Library user. This will resolve issues like the one experienced by Brian Shea, where the agent was unable to generate a response due to lack of access to the knowledge base.
Here are the key permissions required for the Agentforce Data Library user:
By following these steps and granting the necessary permissions, you'll be able to build a robust and effective Agentforce agent that can provide accurate and personalized responses to customer inquiries.
Building an Agent
To create an Agentforce agent, you'll need to follow five key steps, as outlined in the Agentforce process. The first step is to Clarify Roles, defining the agent's purpose, topics, KPIs, and scope.
You might enjoy: Agentforce Consultants
Identify Critical Data is the second step, where you'll determine the data the agent needs to complete its desired actions. You'll also need to ensure your data is complete and ready to use.
Define Instructions & Guardrails is the third step, where you set operational guidelines for your agent by defining core instructions for each topic.
Develop Actions is the fourth step, where you use Apex, Flows, and Prompts to build automations that your agent can call upon to achieve its designated goals.
The final step is to Select Channel, where you expose your Agent to the appropriate channel(s) so it can fulfill its designated purpose.
Here's a brief overview of each step:
By following these steps, you'll be able to create a functional Agentforce agent that can autonomously pivot between distinct tasks to resolve customer service engagements.
Custom Action Using Prompt Builder
You can create a custom action using the Prompt Builder in Agentforce. This approach involves using a Prompt Template Type of "Knowledge Answers" and referencing Resources --> Einstein Search.
A unique perspective: Using Credit Cards in Canada
One of the key benefits of this approach is that it allows you to leverage Einstein Search, which is a powerful tool for searching and retrieving knowledge from your Salesforce org.
To create a custom action using Prompt Builder, you'll need to specify the Prompt Template Type as "Knowledge Answers" and reference the Resources --> Einstein Search. This will enable you to query the Knowledge articles within your Salesforce org.
Here are the key components of a custom action using Prompt Builder:
- Action: Custom Action + Prompt Builder (Prompt Template Type = "Knowledge Answers", Prompt Template instructions references Resources --> Einstein Search)
- Data: Einstein Data Library using Knowledge
Agent Knowledge and Data
Agentforce Data Library is a game-changer for accuracy and customer trust, supporting Knowledge articles, uploaded files, web search, and custom retrievers.
The library enhances Einstein Bots by leveraging large language models (LLMs) to process unstructured data, including chat transcripts, PDFs, audio and video files, and large text files like books.
By harnessing the power of unstructured data, Agentforce can effortlessly process and search this type of information, making it a significant improvement over relying on highly structured data using article answers.
Explore further: Knowledge Process Outsourcing
With Agentforce, you can choose to use an existing Data Library or create a new one, making it easy to get started.
To configure an Agentforce agent, you need to add the General FAQ topic to the agent and assign the standard "Answer Questions with Knowledge" action to the Topic.
A Data Cloud user needs to have specific permission sets, including Data Cloud User, Agentforce Service Agent Object Access, and Service - Knowledge Access, to access Knowledge Articles.
Agentforce uses retrieval augmented generation (RAG) to index Knowledge articles and attachments, ensuring it retrieves the most relevant and up-to-date information.
Here are some key benefits of using Agentforce Data Library:
- Improves agent's response accuracy
- Personalizes answers
- Grounds with retrievers in Data Cloud
To get started with Agentforce, you need to follow these five steps:
1. Clarify roles and define topics, KPIs, and a scope for the agent's actions.
2. Identify critical data and determine whether your data is complete and ready to use.
3. Define instructions and guardrails for the agent.
4. Develop actions using Apex, Flows, and Prompts.
5. Select a channel to expose the Agent to.
By following these steps, you can create an autonomous AI chatbot that can field basic customer questions and schedule service appointments while sounding natural.
Add Topic to Agent
Adding a topic to an agent is a straightforward process, but it's essential to get it right to ensure your agent can provide accurate and helpful responses. If you missed adding the General FAQ topic during the New Agent setup process, you can add it later by clicking "Open in Builder" and adding a Topic with instructions.
To add a Topic, you'll need to assign a standard "Answer Questions with Knowledge" action to the Topic. This will enable your agent to retrieve information from the Knowledge base and provide relevant answers to customer queries.
The "Add Topic to Agent" process is a crucial step in setting up your agent, and it's worth taking the time to get it right. By following these steps, you can ensure your agent is equipped to handle a wide range of customer inquiries and provide accurate and helpful responses.
Here are the steps to add a Topic to an agent:
- Click "Open in Builder"
- Add a Topic and add instructions
- Assign standard “Answer Questions with Knowledge” action to the Topic
By following these steps, you can add a Topic to your agent and ensure it's equipped to provide accurate and helpful responses to customer queries.
The Power of Unstructured Data
Agentforce uses large language models to process unstructured data, which includes chat transcripts, PDFs, audio and video files, legal documents, and large text files like books.
This type of data can be sourced from Knowledge articles or PDF attachments when integrating an Agentforce Data Library in Salesforce.
Unstructured data lacks a consistent format, but large language models can effortlessly process and search this type of information.
Agentforce doesn't need to be trained on how to answer specific questions, as it's built to naturally understand human interactions.
The data library uses grounding with retrieval augmented generation (RAG) to index Knowledge articles and attachments, ensuring it retrieves the most relevant and up-to-date information.
Agentforce accesses the relevant information from the data library and delivers a response that feels natural and conversational when a customer asks a question.
To set up a Data Library, users can choose to use an existing one or create a new one in Agentforce Builder, and then upload attachments or link to existing Knowledge articles.
Data Cloud needs to be activated before setting up a Data Library, and that's all it takes for Agentforce to start working its magic.
Key Differences from Bot Action
Agent Knowledge and Data offers a unique set of features that set it apart from traditional Bot Actions.
The key difference is the ability to have a conversational flow, allowing users to ask follow-up questions and get clarification on specific points.
This is in contrast to Bot Actions, which are limited to a single query and response.
With Agent Knowledge and Data, the system can use the context of the entire conversation to inform its responses, making it more dynamic and user-friendly.
This is particularly useful in scenarios where users need to provide additional information, such as their loyalty level, to get a personalized answer.
For instance, a travel agency's knowledge base might have different answers for Platinum, Gold, and Silver members, and Agent Knowledge and Data can use this information to provide tailored responses.
This level of personalization is not possible with traditional Bot Actions, which are limited to a one-size-fits-all approach.
Here are some key benefits of using Agent Knowledge and Data:
- Ability to ask follow-up questions and get clarification
- Use of context from the entire conversation to inform responses
- Personalized answers based on user information, such as loyalty level
Agent Systems and Strategy
Agent systems can be built in various ways, as seen in the different approaches shared by folks who've worked on similar agents. One approach involves querying unstructured data, such as text, rich text, and PDFs.
The original article sparked a lot of discussion, with many people sharing their own experiences and strategies for building agent systems. One common challenge is finding the right approach for querying unstructured data.
In January 2025, an update was posted, highlighting the different approaches people have taken to build agent systems. This shows that there's no one-size-fits-all solution, and different approaches can be effective depending on the specific needs and goals.
Some people have built agents that query unstructured data, while others have taken a slightly different approach. This diversity of approaches is a testament to the creativity and innovation of those working on agent systems.
The update in January 2025 also mentioned that people have been working on similar agents, with some sharing their approaches and experiences. This community-driven sharing of knowledge and expertise can be a valuable resource for those looking to build their own agent systems.
Broaden your view: Knowledge Sharing
Einstein SDR Agent
Einstein SDR Agents are digital employees that can engage with inbound prospects and nurture pipeline 24/7, generating fully-contextualized responses by drawing upon internal CRM and harmonized external data.
These agents can handle time-consuming sales activities, giving human sales employees more time to build long-term relationships with customers.
They can analyze prospect questions to generate situationally-appropriate answers and recommendations, and autonomously determine common actions such as scheduling meetings or responding to product questions.
Einstein SDR Agents can tackle multiple leads across channels and in various languages, making them a valuable asset for businesses looking to scale their sales efforts.
Here are some key capabilities of Einstein SDR Agents:
- Handle time-consuming, top-of-funnel sales activities
- Analyze prospect questions to generate fully-contextualized, situationally-appropriate answers and recommendations
- Autonomously determine and precipitate common actions (e.g., scheduling meetings, handling low-level complaints or objections, responding to product questions)
Certification and Training
To become a certified agent, you'll need to complete a training program that covers topics like risk management and customer service. These programs are designed to equip you with the skills and knowledge you need to succeed in the field.
Agentforce knowledge is built on a foundation of industry standards, which includes certifications like the Certified Insurance Service Representative (CISR) designation. This certification is a benchmark for excellence in the industry.
To get certified, you'll typically need to complete a training program and pass an exam. The exam will test your knowledge of industry standards and best practices.
Certification Goal
Setting a clear certification goal is crucial to achieving your desired outcome. This goal should be specific, measurable, achievable, relevant, and time-bound (SMART).
It's essential to define what you want to achieve through certification. For instance, you might aim to improve your skills in a particular area or enhance your job prospects.
Research has shown that individuals with a clear goal in mind are more likely to stay motivated and focused throughout the training process. A study found that people with a clear goal in mind were 42% more likely to achieve their desired outcome.
Your certification goal should align with your career aspirations and personal interests. If you're unsure about your goal, consider seeking advice from a career counselor or mentor.
7 Topics
As you dive into the world of Agentforce, you'll want to familiarize yourself with the 7 key topics that make it tick. These topics will help you understand how Agentforce works, how to leverage its features, and how to get the most out of it.
To explain how an agent works and how the reasoning engine powers Agentforce, you'll want to check out the Agentforce Concepts section. This is where you'll learn about the reasoning engine and how it drives Agentforce's decision-making process.
To leverage standard topics, custom topics, standard agent actions, and custom agent actions, you'll want to refer to the Agentforce Concepts section, which outlines these features in detail. This will help you get started with building and customizing your agents.
To manage and monitor agent adoption, you'll want to check out the Agentforce Concepts section, which covers this important aspect of agent management. This will help you keep track of how your agents are being used and make adjustments as needed.
To manage Agentforce user security, you'll want to refer to the Agentforce Concepts section, which provides guidance on user security and permissions. This will help you ensure that your agents are secure and accessible only to authorized users.
To test an agent using the Testing Center, you'll want to check out the Agentforce Concepts section, which outlines the testing process and provides tips for getting the most out of it. This will help you fine-tune your agents and ensure they're working as expected.
To deploy an agent from sandbox to production, you'll want to refer to the Agentforce Concepts section, which covers the deployment process and provides guidance on best practices. This will help you move your agents from testing to production with confidence.
Here's a quick rundown of the 7 key topics to keep in mind as you work with Agentforce:
- How an agent works and how the reasoning engine powers Agentforce
- Leveraging standard topics, custom topics, standard agent actions, and custom agent actions
- Managing and monitoring agent adoption
- Managing Agentforce user security
- Testing an agent using the Testing Center
- Deploying an agent from sandbox to production
Service Cloud 4 Topics
Service Cloud 4 Topics are essential for anyone looking to get certified or trained in this area.
To build an effective Service Cloud 4 solution, you need to know how to build an agent that answers questions based on Knowledge articles. This is a crucial skill for any support or customer service team.
Connecting an agent to a digital channel is also a key aspect of Service Cloud 4. This allows customers to interact with your support team through their preferred channels, such as messaging apps or social media.
Given a scenario, identifying the correct generative AI feature in Agentforce for Service is a critical thinking skill that's essential for any Service Cloud 4 professional. This requires analyzing the situation and choosing the right tool for the job.
Here are some key Service Cloud 4 topics to focus on:
- Building an agent that answers questions based on Knowledge articles
- Connecting an agent to a digital channel
- Identifying the correct generative AI feature in Agentforce for Service
Refreshing and Turbocharging
Gerent is ready to help you take full advantage of Agentforce, promising to redefine how businesses approach day-to-day sales and service operations.
Gerent's dedicated AI & Data practice features talented consultants with a collective wealth of cross-industry, cross-Cloud experience, making them a great partner for Agentforce implementation.
Gerent can deliver seamless, strategic deployment of intelligent agents within your Salesforce environment, enabling you to turbocharge your sales and service operations with confidence.
Here are some key benefits of partnering with Gerent for Agentforce implementation:
- Fast, effective deployment of an Agentforce AI agent tailored to your unique business needs
- Comprehensive knowledge transfer, enabling your team to confidently operate and manage AI agents
- Continuous support to ensure ongoing value as you optimize and refine your AI agents post-deployment
- AI Discovery Workshops + Org Assessment to assess your AI & Data readiness and explore customized opportunities for further automation and efficiency
Refreshing Base
Refreshing Base is a crucial step in keeping your Knowledge Base up-to-date. Agentforce won't immediately have access to changes made to Knowledge articles, so you need to manually refresh the Data Cloud search index.
You might enjoy: Knowledge Base
To refresh the Knowledge Data Stream in Data Cloud, go to Data Stream, find Knowledge_kav_home, and click Refresh Now. The refresh will take a few minutes to run. This step is necessary to ensure your agent can search the updated Knowledge Base.
After refreshing the Knowledge Data Stream, you'll need to rebuild the Search Index in Data Cloud. To do this, go to Search Index, find the Search index associated with the Einstein Data Library you created, and click Rebuild.
Once the Search Index has completed refreshing, your agent should now be able to search the most updated Knowledge Base. You can test this by asking the agent questions that cover your new or updated Knowledge articles and confirming that the agent's responses reflect the most updated versions of the articles.
Here's a quick checklist to help you refresh your Knowledge Base:
- Refresh Knowledge Data Stream in Data Cloud
- Refresh Knowledge Search Index in Data Cloud
- Rebuild the Search Index associated with the Einstein Data Library
By following these steps, you can ensure your Knowledge Base is always up-to-date and your agent is providing the most accurate and helpful responses to your customers.
Turbocharge with General
Turbocharge with Gerent
Gerent can deliver seamless, strategic deployment of intelligent agents within your Salesforce environment. This is thanks to their team of talented consultants with cross-industry, cross-Cloud experience.
Gerent's team can also help you fast-track the deployment of an Agentforce AI agent tailored to your unique business needs. This AI agent is integrated with Salesforce products and set up with scalability in mind.
To ensure a smooth transition, Gerent provides comprehensive knowledge transfer. This enables your team to confidently operate and manage AI agents.
Gerent's support doesn't stop after deployment. They offer continuous support to ensure ongoing value as you optimize and refine your AI agents.
Gerent's services include AI Discovery Workshops + Org Assessment to assess your AI & Data readiness. This helps explore customized opportunities for further automation and efficiency.
Here are some of the key benefits of working with Gerent:
- Seamless deployment of intelligent agents
- Fast deployment of customized AI agents
- Comprehensive knowledge transfer
- Continuous support
- AI Discovery Workshops + Org Assessment
What Is and How To
Agentforce knowledge is built on a foundation of data and analytics, which are used to identify patterns and trends that can inform business decisions. This knowledge is essential for staying ahead in today's fast-paced market.
Data is collected from various sources, including customer interactions, social media, and market research. By analyzing this data, businesses can gain a deeper understanding of their customers' needs and preferences.
To create an effective agentforce knowledge system, businesses should start by identifying their key performance indicators (KPIs). According to the article, KPIs can include metrics such as customer satisfaction, sales growth, and employee engagement.
Once KPIs are established, businesses can begin to collect and analyze data to inform their decision-making. This can be done through the use of data visualization tools, which can help to identify patterns and trends in the data.
Regular training and development programs are also essential for building agentforce knowledge. By providing employees with the skills and knowledge they need to succeed, businesses can improve their overall performance and competitiveness.
A well-designed agentforce knowledge system can have a significant impact on business outcomes. For example, a study cited in the article found that businesses that invested in agentforce knowledge saw a 25% increase in customer satisfaction and a 15% increase in sales growth.
Additional reading: Simple Knowledge Organization System
Featured Images: pexels.com


