Streamline the Student Academic Journey with Agentforce | How I Solved It

Key business problem

High school and college students frequently encounter overwhelming complexity in managing their academic paths. High school students struggle with selecting appropriate schools and managing application materials. College students find course planning, registration, and managing tuition payments complex and confusing. Both student groups face stress and uncertainty due to poorly managed educational processes. Here’s how we identified and are tackling these challenges through Edventure. Our platform ensures that students can make informed decisions with confidence, backed by a system that supports their educational journey every step of the way.

How we solved it

In our college student course enrolling demo, the student engages with the agent to initiate a comprehensive search that aggregates information for each course, including university and course names and course descriptions.

Let’s take a look at this interaction in Agent Builder.

Here, the student inquires about intermediate courses related to models at Harvard. The service agent selects the Academic Planning Topic and invokes the Course Recommendation with University flow agent action.

Let’s take a look at what’s happening in the flow agent action, which is an autolaunched flow.

The flow begins with a (1) robust search for all the courses using the difficulty level, keywords, and university name (all three are inputs from the agent to the flow) and stores the university name, course name, course description, start and end date, and course URL. (2) Then, it determines whether any courses were found. (3) For each course found and identified, (4) the system dynamically adds these courses to a collection record variable. This allows for an organized compilation of all relevant courses, so the agent can be ready for the next recommended action on course registration.

When the student asks to enroll in the course, the agent asks for additional information (first name, last name, and email address) to validate the student’s identification per the instruction in the Academic Planning topic, as we see in the planner in Agent Builder.

Once the student provides the information, the agent selects the Academic Planning topic and invokes the Registrant Add Student flow action.

Agentforce streamlines the course registration process within the student system using a flow agent action Registrant: Add Student autolaunched flow.

1. We use a Get Records element to find the student record using email, first name, and last name—details provided through the agent's user prompt.
2. We use another Get Records element to retrieve existing registrants by querying the student’s ID and course name, filtering out any records with a status of canceled or failed.
3. Next, we use a Decision element to verify whether the student is already registered for the course.
4. Assuming the student is not already registered, we search for the course using the provided course name, ensuring it starts on a future date and is associated with the specified university with a Get Records element.
5. We create a new registrant record for the student using a Create Records element.
6. The newly created registrant record is retrieved as a verification step with a Get Records element.
7. Finally, using an Assignment element, the verified student’s details are assigned to a record variable, which the agent can then utilize.

Our second scenario involves a high school student applying to Cornell University. They use an agent to streamline the procedure to handle university match. But before the agent can perform the action, it asks for student verification using contact information.

In Agent Builder, we can see the agent selects the Academic Planning topic and uses two of its instructions to determine next steps.

Once the verification information is provided, the agent gives the probability of the student being selected as a marketing major at Cornell. This was done by selecting the Academic Planning topic and invoking a Verify Create Student Contact flow agent action proceeded by a Recommend University flow agent action.

Let’s take a look at the first flow action, Verify/Create Student-Contact.

1. We search for the student's record with a Get Records element using the email, first name, and last name. These details are provided via a user prompt and passed into the flow to locate the existing contact.
2. Using a Decision element, we determine if the student's contact information is found. The flow proceeds to the next steps. This determination ensures that existing records are identified before any new entries are created.
3. If the student does not already exist, we create a new student contact record using a Create Records element.
4. We verify the contact record with a Get Records element, retrieving the record to confirm that the new student contact now exists in Salesforce.
5. Once verified, the newly created student’s details are assigned to a record variable with an Assignment element. This variable is then passed back to the agent, completing the flow and ensuring the student's information is ready for further processing or interaction.

Once the student is verified, it triggers a flow that performs the following steps to recommend suitable university majors.

1. We retrieve the student's contact record using the email provided with a Get Records element. It ensures that essential academic information—such as GRE Score, TOEFL Score, Statement of Purpose (SOP), Letters of Recommendation (LOR), and Cumulative Grade Point Average (CGPA)—is all present and accounted for.
2. Using a Decision element, we determine if a student record is found with these qualifications.
3. If a student record is found, we use a Get Records element to search for majors that align with the student’s interests and academic qualifications, using major and university details specified by the agent.
4. We use another Decision element to determine whether relevant records for the major have been found.
5. If records were found, for each major identified...
6. ... we use a Decision element to evaluate whether the student meets the specific qualifications required for that major.
7. If the student qualifies for the major, we add the major to a collection variable with an Assignment element. This collection of qualified majors is then prepared to be sent back to the agent for further action.

And that’s it! Two different use cases solved by our Edventure agent!

Business results

Despite being in its initial testing phase, the results from Edventure have been profoundly encouraging. This system is designed to not only streamline the academic planning process but also enhance the match between students and their ideal academic programs and institutions. In Agentforce, we added customized agent actions using flows to handle tasks such as university locations, rankings, professor rating, etc. As we continue to refine Edventure, our goal remains steadfast: to empower students to navigate their academic journeys with ease and confidence.

Do try this at home

At the core of our Salesforce configuration, flow agent actions power the service agent, and mastering this can greatly enhance your business creativity and productivity. Using Trailhead, you can explore and practice creating your own business flows, gaining practical experience with a tool that’s indispensable in today’s CRM landscape.

Moreover, don’t limit the application of these flows to just complex scenarios. Try implementing similar flows for more routine tasks in your Salesforce environment. For example, our process of verifying student information and recommending courses or majors can be adapted to other areas such as customer service inquiries or sales lead assignments. By extending this logic to other standard objects within Salesforce, you can solve a variety of business problems with similar strategies.

Resources

• Salesforce Admins Blog: Harness the Power of AI, Flow, and Slack to Summarize and Notify Users
• Trailhead: Quick Start: Build Your First Agent with Agentforce
• Salesforce Help: Using Retrieval Augmented Generation
• Salesforce Developers site: Agentforce Workshop | Build the Future with Agentforce
• Salesforce Developers site: Agentforce Developer Guide | Agentforce and Einstein Generative AI
• Salesforce Developers site: Data Cloud Documentation

Want to see more good stuff? Subscribe to our channel!