How PNC Learning resolves 65% of student support requests with Hiver AI

Founded in 2017, PNC Learning provides accredited continuing education and licensing exam preparation for CAIB*, Fundamentals of Insurance, RIBO, LLQP, Alberta, or Texas P&C insurance exams. The company operates across Canada and is expanding into the US, with a globally distributed customer support team.

Supporting students in regulated industries comes with unique challenges. Licensing requirements vary by region, exam registration processes can be complicated, and students often need help navigating multiple steps before earning their certifications.

Today, PNC Learning uses Hiver’s AI Agents to resolve 65% of student support requests, while its team focuses on conversations that need personal attention.

The challenge: Resolving increasingly complex student support at scale

As PNC Learning grew, its support team faced an increasing number of questions about course pricing, enrollment, licensing requirements, and exam registration.

Many of these requests were repetitive, but answering them correctly required understanding different regional regulations, course requirements, and certification processes.

PNC initially experimented with Zapier and GPT-based automation, but faced two challenges:

• High automation costs: Zapier’s task-based pricing made running support workflows expensive.
• Manual AI maintenance: The team had to review conversations, update prompts, and repeatedly feed new information into GPT to keep responses accurate.

PNC needed AI Agents that could independently resolve student questions, access information across systems, and improve over time without constant manual intervention.

How PNC Learning deployed Hiver’s AI Agents to resolve 65% of requests

1. Training AI Agents to handle student questions end to end

PNC Learning started by training Hiver’s AI Agents on the same standard operating procedures, product guides, and onboarding documents used to train its support team. They uploaded their internal documents and connected their website content to give the agents the knowledge they needed to answer student questions.

Accuracy was particularly important because regulatory information isn’t always straightforward. Before going live, the team tested how the AI Agents responded to different student queries, refined their knowledge, and built confidence in the accuracy of their answers.

Once deployed on PNC Learning’s website, the AI Agents began independently handling questions such as:

• What does a particular course cost?
• Which course is required for a specific license?
• How do I register for my licensing exam?
• What steps do I need to complete to get certified?

The team also drew a clear boundary between AI and human support. AI handles transactional questions, while people step in for refund requests, students who have failed exams, and conversations requiring empathy or judgment.

2. Connecting AI Agents to Thinkific to resolve more complex requests

PNC Learning wanted its AI Agents to do more than answer questions using FAQs. Many student inquiries require information from Thinkific, the company’s learning management, billing, and student records platform.

According to Jack, more than 90% of the information needed for student support lives in Thinkific.

Working with Hiver, PNC built a custom Thinkific Skill that allows its AI Agent to retrieve course information through APIs and use it to respond to students.

For example, when a student asks about a course or available discount, the agent can:

1. Understand the request: Identify the course, pricing, or discount information the student needs.
2. Ask follow-up questions: Determine the student’s region when course availability or selection depends on location.
3. Retrieve information: Call the appropriate Thinkific API to find the relevant course details.
4. Respond to the student: Use the retrieved information to provide an accurate, contextual answer.

This is particularly useful for PNC because different regions have different courses and certification requirements.

The Thinkific Skill extends AI beyond static knowledge to information stored in another business system. PNC is also exploring student-specific lookups, including enrollment status and course expiration dates, which still require manual intervention.

3. Continuously improving AI responses using resolved conversations

One of PNC’s biggest challenges with its earlier GPT workflows was keeping AI knowledge accurate. Every time something changed, someone had to review conversations, identify gaps, and manually update prompts.

With Hiver Improve, PNC can identify knowledge gaps from actual support conversations and use those insights to improve its AI Agents.

One example involved a course where PNC delivered part of the training, while an external provider handled the certification required by the regulator. Explaining the process was complicated enough that even human representatives initially needed additional training.

Hiver Improve surfaced conversations where the AI Agent needed more information to explain the process. The team corrected the gap, improving how the agent handled similar questions.

Instead of manually sampling conversations, Jack now reviews Improve periodically to see where the agent needs attention. This makes it easier to keep responses current as courses, processes, and student questions evolve.

The results: 65% of student requests resolved by AI Agents

With Hiver AI Agents, PNC Learning has automated a significant portion of student support while giving its team more time for situations that need human involvement.

• 65% of student requests resolved by AI: Agents independently handle recurring questions about courses, pricing, licensing, and registration.
• Complex requests handled through connected tools: The Thinkific Skill retrieves course information through APIs, allowing agents to answer questions using data from business systems.
• Less effort maintaining AI knowledge: Hiver Improve identifies gaps from real conversations, reducing the need for manual ticket reviews and prompt updates.
• More time for high-touch support: Human representatives focus on sensitive conversations, refunds, and students who need personalized guidance.

For PNC Learning, AI is becoming a core part of how the company runs student support, resolving requests, learning from real conversations, and working across connected systems, while its team focuses on students who need personal attention.