Configuring Sales Assistant in Pega Sales Automation
5 Tasks
15 mins
Beginner
English
Scenario
Acme Technologies wants to improve sales team productivity by configuring a Sales Assistant agent in Pega Sales Automation™. The agent will support sales representatives in managing opportunity data, scheduling meetings, and automating tasks.
As a sales administrator, your task is to configure the Sales Assistant agent with the appropriate instructions and context, so that it can analyze emails, appointments, tasks, activities, opportunities, leads, and contacts. The agent must provide a list of the top five intelligent recommendations, each accompanied by clear reasoning. Additionally, it must respond accurately and perform actions that enhance the productivity of the sales team.
The following table provides the credentials you need to complete the challenge:
| Role | User name | Password |
|---|---|---|
| Sales Administrator | Sfasamplesysadmin | rules |
| Sales Representative | Tmason | rules |
Your assignment consists of the following tasks:
Task 1: Create an agent
Create a Prioritization Actions agent in the Generative AI section.
Task 2: Define agent instructions, guardrails, and tone
On the DEFINITION tab, provide agent instructions, guardrails, and tone to enhance user interactions with the agent.
Task 3: Set up context Data Pages
In the Additional Context section, list the Data Pages that the agent requires.
Task 4: Add advanced tools and chaining
On the ADVANCED TOOLS tab, add actions the agent can perform.
Task 5: Select AI model
On the SETTINGS tab, select the AI model for the agent.
Task 6: Confirm your work
Test the scenario by logging in as a sales representative. Begin the interaction by asking Act as a prioritization agent and list the top 5 actions to perform across the provided sales automation work objects.
The agent analyzes emails, appointments, tasks, activities, opportunities, leads, and contacts to provide intelligent recommendations. It recommends a list of top 5 action items with clear reasoning.
Challenge Walkthrough
Detailed Tasks
1 Create an agent
- In the Pega instance for the challenge, click Launch Pega Infinity to log in to Infinity Studio:
- Click App > Records.
- Click Generative AI > Agent.
- Click the Create button to create a new agent.
- In the Label field, enter Prioritization Actions.
- In the Context section, verify that the following details are configured:
- Application layer: Sales Automation
-
Apply to: PegaCRM-Portal
-
Add to ruleset: SAPlusC11N
-
Version: 01-02-03
- Click Create and Open.
2 Define Agent instructions, guardrails, and tone
- Download the sample Agent Instructions, guardrails, and tone file for the Prioritization Actions agent to define the agent’s instructions, role, guardrails, expected behavior, response tone, and how it should answer user questions.
- On the DEFINITION tab, in the Instructions field, define the context and expected responses.
For example, include work object definitions such as opportunities, leads, appointments, and emails. Specify how the agent should respond to user queries and whether the data should be loaded at agent initialization or on-demand. - In the Guardrails section, define validation rules to ensure that user inputs are in the correct format.
For example, validate that email IDs are in a proper format. - In the Response Style & Tone section, specify the tone of the agent's responses.
Apply soft, casual, or formal tones depending on your use case. For example, use language that is empathetic and solution-oriented, avoiding jargon unless contextually appropriate.
3 Set up context Data Pages
In the Additional Context section, click Add Data Page to list the Data Pages that the agent requires.
For example, include data pages for the following items:
|
Name |
Data Page |
|
Leads |
D_FetchLeadsForCoach |
|
Opportunities |
D_FetchOpportunitiesForCoach |
|
Contacts |
D_FetchContactsForCoach |
|
Recent Emails and Appointments |
D_FetchRecentAppointmentsAndEmails |
|
Accounts |
D_FetchAccountsForCoach |
|
Open and Completed tasks |
D_FetchTasksForCoach |
|
Actions from completed and upcoming appointments |
D_FetchActionsForCoach |
Note: If the agent needs all data at load time, include the Data Pages in the Additional Context section. If the agent should fetch data only when needed, configure the Data Pages on the Knowledge tab.
4 Add advanced tools and chaining
- On the ADVANCED TOOLS tab, click Add to define the actions the agent can perform.
- In the Tool field, enter ScheduleMeeting, and then press the Enter key.
- Repeat steps 1-2 to add the CreateTask and SendEmail actions.
Note: To enable agent chaining, specify any child agents that the parent agent should invoke.
5 Select AI model
- On the SETTINGS tab, in the Model section, select the AI model that you want the agent to use.
In this example, select the GPT-4o-Mini model for better results.
- On the HISTORY tab, enter the following description:
This sales assistant agent is used to suggest/recommend action items and run them and search the data/update the opportunity info etc. This agent helps the sales reps/sales managers on closing the deals successfully. - Click Save to save the configuration.
- In the navigation pane, click Channels > Channels Landing Page.
- On the Current channel interfaces landing page, click SA Portal.
-
Scroll down to the Agents section, click Add.
-
Click Prioritization Actions, and then click Submit.
-
Click Save, and then Log off.
Available in the following mission:
If you are having problems with your training, please review the Pega Academy Support FAQs.
Want to help us improve this content?