This is a cache of https://developer.ibm.com/articles/prebuilt-domain-agents-watsonx-orchestrate/. It is a snapshot of the page as it appeared on 2025-11-26T06:46:38.494+0000.
Building and deploying prebuilt domain agents in watsonx Orchestrate - IBM Developer
In this article, we’ll explore the robust library of prebuilt domain agents in IBM watsonx Orchestrate, showcasing its Agent Catalog and diving deeper into pre-built domain agents for HR agents with SAP SuccessFactors and Sales agents with Salesforce, Dun & Bradstreet, Seismic, and Outlook.
These ready-to-deploy agents offer significant business advantages: rapid deployment with enterprise-grade, domain-specific logic, centralized governance and observability, and plug-and-play integrations with existing systems. For enterprise developers, they accelerate solution delivery by reducing the need for custom coding, provide reusable components that align with best practices, and simplify integration through standardized agents, tools, and prebuilt connectors. With these prebuilt domain agents, developers can focus on innovation and higher-value tasks while ensuring scalability, security, and compliance through intelligent automation.
The watsonx Orchestrate Agent Catalog
The IBM watsonx Orchestrate Agent Catalog provides a diverse selection of prebuilt domain agents across functions like HR, sales, customer care, finance, IT, and more.
As of this writing, the catalog features 232 prebuilt domain agents and 831 prebuilt tools, each capable of performing specific actions, retrieving information, or complex reasoning. This rich ecosystem is expanding rapidly, delivering ongoing access to new capabilities and enterprise applications.
These prebuilt agents are designed to function out-of-the-box with a broad range of applications, facilitating automation across essential business platforms. They integrate natively with systems such as Coupa, Dropbox, Dun & Bradstreet, Salesforce, Outlook, Gmail, ServiceNow, and SAP SuccessFactors, among others. Enterprises can capitalize on present investments without extensive custom development. Whether managing files in Dropbox, updating CRM records in Salesforce, managing IT service requests in ServiceNow, or synchronizing HR data in SuccessFactors, these agents serve as intelligent bridges, unifying workflows, reducing data inconsistencies, eliminating silos, and lowering manual effort and operational complexity.
Building and deploying prebuilt HR agents
In this guided walkthrough, you will explore the steps that are required to use the prebuilt HR domain agents for SAP SuccessFactors, starting from the configuring the connections to SAP SuccessFactors, deploying the agent, and then testing it with business scenarios.
In this example, we use the Employee Support Manager (SAP SuccessFactors) agent that oversees employee-related HR queries. Acting as a central coordinator, it interprets user requests and seamlessly routes them to most suitable specialized support agent.
Configure secure connections between the prebuilt HR agents and the required enterprise applications
The prebuilt domain agents require secure connections to the third-party applications that they depend on. To configure this secure connection to your enterprise application, follow the IBM Documentation for your enterprise application.
Employee Support Manager (SAP SuccessFactor) agent requires connection to SAP SuccessFactors. The process typically involves configuring the connection to your application, which includes configuring authentication, providing access to the API, and saving the connection.
Build and deploy prebuilt HR agents
In this section, you will learn how to use the Employee Support Manager agent for SAP SuccessFactors as a template and customize it for your HR workflows.
Important: This article is simply a guided walkthrough of how to build and deploy prebuilt domain agents with enterprise applications. Do not try to follow these steps, even if you have the SAP SuccessFactors application, but instead review the steps and screen captures to gain an understanding of how you will work with the prebuilt domain agents overall.
To build and deploy prebuilt HR agents with SAP SuccessFactors, you’ll need to:
Filter the Agent Catalog by HR and SAP SuccessFactors.
Identify the Employee Support Manager supervisor agent.
Explore the supervisor agent profile and the collaborator agents.
Modify the agent name and review its behavior.
Apply the recommended reasoning model.
Review the available channels.
Deploy the customized agent in watsonx Orchestrate.
Let’s walk through these steps now.
In the Agent Catalog, filter by:
Type: Agents
Categories: HR
Apps: SAP SuccessFactors
You will see list of domain agents handling compensation, personal details, visa, and mobility.
Search for “Employee Support Manager” and then select it.
This is the supervisor agent within SAP SuccessFactors that oversees employee-related HR queries. It acts as a central coordinator, interprets user requests, and seamlessly routes them to most suitable specialized HR support agent. The following screenshot shows an example of the specialized agents that this supervisor agent uses, along with high-level description of each agent.
To use this prebuilt agent, you click the Use as template button.
Optionally, you can update its name by clicking the pencil icon.
The agent profile provides the purpose of this agent. The Employee Support Manager is a supervisor agent in SAP SuccessFactors that centralizes HR queries and routes them to specialized agents, streamlining tasks like personal data updates, benefits access, internal mobility, leave requests, and organizational changes for faster, more accurate HR workflows.
In the Toolset section, in the Agents section, to view the collaborator agents. The Employee Support Manager delegates tasks to specialized agents such as Compensation (4 tools), Employee Email (3 tools), Employee Visa (2 tools), Employee Personal Info (6 tools), Employee Phone, and Employee Address, each equipped with dedicated tools to handle specific HR actions efficiently.
In the Behavior section, you can see how the agent responds to different types of user requests. Since this is a template, you can adjust the description and other behavior settings to better fit your use case when needed.
IBM watsonx Orchestrate includes a huge set of supported models including models from Meta, Groq, and others. You can change the reasoning model from the Model menu.
You can also make the agent available across different channels including embedding it in a chat UI, Microsoft Teams, WhatsApp, Genesys, Slack, and others.
To deploy the agent, click Deploy.
It provides a pre-deployment summary including agent profile, knowledge, tools, collaborator agents, and behavior. It also shows that the connection to SAP SuccessFactors is configured. Review and confirm by clicking Deploy.
test prebuilt HR agents with SAP SuccessFactors
In this section, you will experience how the Employee Support Manager HR agent actually works. You do not need to have access to a SuccessFactors environment, but you can simply review the steps and screen captures to understand the context and see how prebuilt agents handle real-world HR queries and actions.
You navigate to the Chat view in watsonx Orchestrate.
In the Agents drop-down, you click “SuccessFactors Employee Support Manager”.
If you ask the agent, “Show me payslips for jadannb@test.com for 2018-01-01 to 2018-01-31,”check out the reasoning.
If you click Show Reasoning, you can explore how the agent retrieved the results. In this example, the Employee Support Manager agent identified that the request is related to compensation, so it delegated the task to the Compensation agent. The agent then used a prebuilt tool called “get_user_id_in_sap_success_factors” to get the user ID of jadannb@test.com, and then called another tool called “get_payslip_details_in_sap_success_factors” to retrieve the payslip information.
The agent remembers the context of the conversation, so you can ask the agent follow-up questions related to the same employee: “Show me the current compensation details”.
You can review the compensation details for the same employee on SuccessFactors to validate the agent logic again.
Note: This is not an actual person, and it is a test data for illustration purpose.
Agents are not designed to retrieve information only. They also help in executing actions. Ask the agent to update her phone number.
Observe the reasoning. The agent transfers the request to another specialized agent called employee_phone to handle the request.
The power of agent reasoning appears when dealing with complex queries.
For example, checking whose visa is expiring soon in a department, and then you can even ask the agent to take an action based on that, send an email or even start visa renewal process.
As you can see dynamic reasoning happens in 9 different steps decided by the agent to check which employee has a visa expiring soon including powerful multi-agent orchestration to the organization_structure agent, employee_visa agent, and Employee Support Manager agent.
In the Manage agents section, you can update its behavior so that it responds only in Spanish.
Building and deploying prebuilt Sales Prospecting agents
In this guided walkthrough, you will explore the steps required to use the prebuilt Sales prospecting agent, starting from the configuring the connections to the required enterprise applications, deploying the agent, and then testing it with business scenarios.
You will be using the Sales Prospecting agent is designed to coordinate sales-related tasks by integrating with key enterprise systems. It connects to Salesforce for managing accounts, leads, and opportunities; Seismic for accessing product collateral and enablement content; and Dun & Bradstreet for enriched firmographic, financial, and contact data. These integrations allow the agent to provide account insights, retrieve industry and competitive information, and support personalized outreach through Outlook. By orchestrating these capabilities, the agent helps sales teams research prospects, access relevant content, and execute CRM updates efficiently.
The Sales Prospecting Agent delegates tasks to specialized collaborator agents, each integrated with a specific system:
Activity Management on Salesforce: Streamlines activities like creating, viewing, and modifying tasks and events within Salesforce.
CRM Management for Salesforce: Handles customer relationship management tasks, including account, lead, and opportunity updates.
Client Outreach with Outlook: Connects with verified clients and supports personalized email communication through Outlook.
Product Enablement with Seismic: Retrieves product collateral and enablement content from Seismic for accurate product information.
Client Search and Enrich with Dun & Bradstreet: Enhances prospecting by leveraging enriched company data for better targeting.
Sales Research with Dun & Bradstreet: Provides client and industry research powered by, including firmographic and financial insights.
Configure secure connections between the prebuilt Sales Prospecting agents and the required enterprise applications
The Employee Support Manager (SAP SuccessFactor) agent requires connection to Salesforce, Dun & Bradstreet, Seismic, and Outlook. The process typically involves configuring the connection to your application including configuring authentication, access to API, and saving the connection.
You can configure connections to your enterprise applications through following IBM documentation namely:
Build and deploy prebuilt Sales Prospecting agents
In this section, you will learn how to use the Sales Prospecting agent with Salesforce, Seismic, Dun & Bradstreet, and Outlook as a template and customize it for your sales prospecting workflows.
Important: This article is simply a guided walkthrough of how to build and deploy prebuilt domain agents with enterprise applications. Do not try to follow these steps, even if you have the Salesforce, Seismic, Dun & Bradstreet, and Outlook applications, but instead review the steps and screen captures to gain an understanding of how you will work with the prebuilt domain agents overall.
Let’s walk through these steps now.
We need to search for “Sales Prospecting Agent” in the watsonx Orchestrate Agent Catalog, and click Use as template.
You can optionally rename the agent to include your company name, review its collaborator agents, and then click on Deploy to deploy the agent.
For simplicity, we renamed it to “Sales Prospecting Super Agent”, and it’s a supervisor agent that delegates tasks to several specialized agents.
The pre-deployment window confirms that all required connections are set properly, so you can just click Deploy.
test prebuilt Sales Prospecting Agents
In this section, you will experience how the Sales Prospecting agent actually works. You do not need to have access to a Salesforce, Seismic, Dun & Bradstreet, or Outlook environments, but you can simply review the steps and screen captures to understand the context and see how prebuilt agents handle real-world sales prospecting queries and actions.
All details in this scenario are for illustration purposes only and use sandbox, non-production environments of Salesforce, Seismic, Dun & Bradstreet, and Outlook. Any company names appearing in examples are illustrative and doesn’t represent real or accurate information.
To start the conversation, you go to the agent chat and select the “Sales Prospecting Super” agent from the drop-down list.
This example demonstrates how the Sales Prospecting Agent helps identify potential clients in a specific region and industry.
For example, you can ask, “Find 10 clients in electronics in California.” The agent interprets the request and returns a curated list of companies in the electronics industry within California, including details such as employee count and yearly revenue.
Behind the scenes, the agent uses multi-agent orchestration to fulfil this request.
The agent first delegates the query to the Client Search and Enrich agent.
Then, it calls the Dun & Bradstreet tools to retrieve details about the clients and their financial data.
This ensures accurate, enriched insights for sales teams without manual research.
Using the Sales Research with Dun & Bradstreet agent, you can retrieve recent news and industry updates about a specific company to enrich prospecting conversations.
For example, you can ask, “Fetch me recent news about a specific company”.
The agent delegates the request to the Sales Research collaborator agent, that uses the get_news_and_media_from_dun_bradstreet tool to return curated articles from trusted sources.
These include insights on partnerships, product developments, and market trends, helping sales teams stay informed and tailor outreach with relevant context.
Using the Product Enablement with Seismic agent, you can access tailored product collateral and messaging to support personalized outreach.
For example, you can ask, “How can I pitch watsonx to a marketing persona in this company?”
The agent delegates the request to the Product Enablement collaborator agent, which uses the product_enablement tool to retrieve relevant enablement content from Seismic. It returns a structured response including a title, introduction, and problem statement crafted for the marketing persona. Importantly, the agent remembers the context from previous steps, such as the company’s industry and focus areas, ensuring that the pitch aligns with prior research and prospect details.
Using LLM capabilities combined with the Client Outreach agent that is integrated with Outlook, you can draft and send personalized emails based on the context gathered in previous steps.
For example, you can ask, “Draft an email to address explaining the above from person.” The agent generates a professional email tailored to the company’s industry and prior research. It leverages the context from earlier steps to ensure relevance and personalization. Once reviewed, the email can be sent directly through Outlook, completing the outreach workflow seamlessly.
Using the CRM Management agent integrated with Salesforce, you can create leads directly from the conversation.
For example, you can ask, “Add this lead on Salesforce.”
The agent uses the create_lead_in_salesforce tool to add the lead with all relevant details like company name, revenue, location, industry, and contact information. This eliminates manual data entry and ensures accurate CRM updates in real time.
You can also retrieve account and opportunity details for the same industry from Salesforce using the CRM Management agent.
For example, you can ask, “Show me 3 accounts on Salesforce related to the same electronics industry.” The agent uses the list_accounts_in_salesforce tool to return a list of accounts in the same industry.
Next, you can ask, “Show me all opportunities in a specific client.”
The agent then uses the list_opportunities_in_salesforce tool to display all opportunities linked to that account, including amount, close date, probability, and stage name.
You can also update activities using the Activity Management agent on Salesforce, which streamlines tasks such as creating, viewing, and modifying activities and events directly within Salesforce helping sales teams maintain accurate engagement records without leaving the workflow.
This guided walkthrough demonstrated how Sales Prospecting prebuilt domain agents in watsonx Orchestrate enable sellers to seamlessly manage the entire prospecting lifecycle from identifying the new leads, and doing sales research with Dun & Bradstreet, to finding product enablement materials in Seismic, and finally managing the leads, accounts, opportunities, and activities within Salesforce.
Summary and next steps
In this article, we introduced you to the watsonx Orchestrate Agent Catalog, which includes hundreds of prebuilt agents and hundreds more prebuilt tools across numerous domains, including HR, Sales, Finance, IT, and more.
Then, we provided a guided walkthrough of the steps you need to take to build and deploy the prebuilt domain agents: configuring secure connections, customizing the prebuilt agent as a template and deploying it, and then testing the agent.
The value of prebuilt domain agents lies in their ability to combine structured automation with intelligent reasoning, enabling organizations to orchestrate complex, multi-step processes without coding. They reduce manual effort, improve accuracy, and accelerate solution delivery while maintaining transparency and control.
As your next steps, explore more of the prebuilt agents in the Agent Catalog, try out the low-code (no-code) features in watsonx Orchestrate, and try out developing your own agents with the ADK. Also, check the watsonx Orchestrate documentation and IBM Developer tutorials for more on how to best use watsonx Orchestrate in your environment.
Acknowledgments
This tutorial was produced as part of the IBM Open Innovation Community initiative: Agentic AI (AI for Developers and Ecosystem).
The authors deeply appreciate the support of Ela Dixit, and Michelle Corbin for the guidance on reviewing and contributing to this tutorial.
About cookies on this siteOur websites require some cookies to function properly (required). In addition, other cookies may be used with your consent to analyze site usage, improve the user experience and for advertising.For more information, please review your cookie preferences options. By visiting our website, you agree to our processing of information as described in IBM’sprivacy statement. To provide a smooth navigation, your cookie preferences will be shared across the IBM web domains listed here.