How a Healthcare Provider Used Stack AI to Speed Up Customer Support

In the following case study, we’ll examine how a virtual healthcare provider used Stack AI to improve efficiency and quality of care.

The healthcare provider offers primary, specialty, and urgent care. Patients can book telehealth appointments with certified physicians, order prescriptions, perform lab tests, and much more. The healthcare provider currently partners with thousands of licensed medical professionals across the United States.

The Opportunity

The healthcare provider serves a wide volume and variety of patients, each with their own unique medical needs. The healthcare provider also fields a massive amount of patient phone calls and messages. These burdens make customer support a more difficult task than for a standard consumer company.

Additionally, the healthcare provider must share accurate healthcare information with patients at all times. If the healthcare provider divulges inaccurate information, these errors can lead to adverse healthcare outcomes. This can cause negative patient experiences, and potential legal liability.

Furthermore, customer support in the healthcare industry must adhere to strict privacy and security guarantees. This includes HIPAA and other protocols for protecting PII and network security.

With the volume of patient inquiries increasing daily, the healthcare provider needed to expand its customer support operation. But the provider had to meet the stringent requirements of healthcare, while also keeping costs low.

After researching solutions, the healthcare provider wanted to use AI agents to automate customer support. But using AI agents in enterprise organizations, especially in healthcare, came with its own challenges. That’s why the provider turned to Stack AI, an enterprise-grade solution for deploying AI agents.

The Solution

With Stack AI, technical and non-technical teams at the healthcare provider built AI agents using a no-code, drag-and-drop interface. Stack AI allowed business teams with no coding experience to create AI agents from pre-built components. This allowed teams to launch AI agents in less than a day, as opposed to handing off the project to overburdened developers, and waiting weeks or months in the queue.

The healthcare provider created an AI agent that could help customer support representatives assist patients. The AI agent allowed customer support reps to enter a patient’s questions about health issues, insurance, scheduling, and other medical topics. The agent used the patient ID to retrieve the patient’s medical history.

The Anthropic LLM model answered the question based on the patient’s medical history, as well as the healthcare provider’s official healthcare documentation. Although the AI agent did not require coding to build, the team did incorporate snippets of Python to access patient data from the healthcare provider’s databases. Finally, the LLMs were configured with detailed prompts to respond like a medical professional.

This AI agent helped customer support reps answer patient questions faster and more efficiently. By retrieving the patient’s medical history, the AI agent could base its answers on the patient’s background, leading to accurate medical advice. Once the AI agent was deployed, customer support reps could assist more patients each day, resulting in shorter backlogs and wait times.

Unlike many builder tools, Stack AI is built for enterprise organizations that require strict privacy and security. With Stack AI, the healthcare provider could build AI agents while complying with all the necessary security protocols in healthcare, including HIPAA. This allowed the healthcare provider’s AI agent to handle the sensitive medical information of patients. The healthcare provider was also able to host Stack AI on premise, enabling enhanced security and data privacy.

With the success of the customer support AI agent, the healthcare provider began to create more AI agents for jobs across the organization. The healthcare provider created an AI agent that allowed physicians to quickly access a patient’s medical history. Instead of rifling through various systems, physicians could now access all the patient information they needed directly from the AI agent.

Another AI agent designed by the healthcare provider included a call center message classifier. The agent listens to a recorded phone call and judges if the call complies with regulation and protocols. This ensures that customer support reps are adhering to all internal and external policies.

The Results

When the healthcare provider came to Stack AI, the organization needed a way to speed up support calls for patients. With Stack AI, the healthcare provider developed AI agents for customer support that answered patient questions quickly and accurately while still complying with healthcare regulations.

After this initial success, the healthcare provider began using AI agents in many different contexts, making medical professionals more effective. AI agents have improved the patient experience, and the healthcare provider continues to build new agents to realize the benefits of generative AI.

2.5 million

AI Agents executions

475,000

Hours saved

12 seconds{' '}

Support wait time