Amid rising healthcare burnout, healow wanted to develop a solution that would use AI to free up staff and handle patients’ calls.
healow created the contact center solution Genie, which will use Voice Live API in Azure AI Foundry to provide patients with an AI-powered experience when they call in.
A seamless, low-latency experience allows physicians to answer more calls from patients, reduce administrative work, and keep satisfaction high.
In healthcare, patient experience is always the top priority. But as burnout among healthcare workers continues to rise and shortfalls in staff challenge traditional operations, it’s safe to say the industry is looking for ways to modernize while keeping patients at the center of care. healow—a comprehensive health IT platform offering a full range of patient engagement and workflow optimization solutions—has been modernizing operations for practices, patients, hospitals, health systems, and payers for the past 10 years. And the company’s healow Genie contact center solution is now taking that modernization a step further.
Healthcare providers like hospitals and doctor’s offices that plan to use Genie can offer an AI-driven medical receptionist for patients who call in. healow is piloting Voice Live API, a feature of Azure AI Speech in Azure AI Foundry for Genie to inform patients about their upcoming appointments, answers common questions, and returns voicemails. Reducing these routine calls saves healthcare staff hours each day and boosts patient satisfaction through timely interactions.
“Genie is offering provider practices a way to use AI to converse with patients through their preferred channel. This will reduce the amount of administrative work and cost for practices to simply give patients the answers to their questions.”
Sidd Shah, Vice President of Strategy & Business Growth, healow
The ways in which doctors communicate with patients have evolved over the years. What used to involve mostly calls and faxes, now includes a mix of patient portals, SMS, chatbots—and yes, calls as well. But reaching out over the phone is resource-intensive for healthcare practices. It usually requires an administrative staff member to call patients and take the time to answer questions. Sometimes, this staff member makes dozens of calls each day to patients who are all receiving the same procedure, so it becomes very routine. Dedicating resources to these tasks also impacts budget, yet it doesn’t necessarily boost patient satisfaction. For instance, if a patient receives a call but doesn’t answer, the practice staff will likely leave a voicemail. However, if the patient calls back outside of working hours, it might kick off a long game of phone tag that ends in frustration.
“Patients calling through voice channels is still one of the most common, popular ways to reach out to physicians,” said Sidd Shah Vice President of Strategy & Business Growth at healow. “But it takes up a lot of resources from practices. And as practices grow, they need more resources to keep up with patient demand and they likely won’t be able to support all the incoming calls they get. This leads to patient dissatisfaction because they’re not getting a call back or they’re not getting answers to their questions.”
As an AI-powered contact center solution, Genie was created to handle patient communications and free up healthcare staff. And now its voice capabilities will take those communications to the next level. “Genie is offering providers a way to use AI to converse with patients through their preferred channel,” Shah said. “This will reduce the amount of administrative work and cost for practices simply to give patients the answers to their questions.”
For Genie to be effective, the healow team needed to build it with AI solutions that offered real-time and natural language capabilities, as well as global scale and enterprise-grade security. They chose Voice Live API in Azure AI Foundry because it offers unified, real-time speech-to-speech capabilities. Voice Live API integrates speech-to-text (STT), generative AI models, text-to-speech (TTS), avatar, and conversational enhancement features into a single, low-latency pipeline, so the healow team didn’t have to stitch together disparate components. “We’re looking forward to using Voice Live API so that when a patient calls, Genie can detect the question and respond in a natural voice in near real time,” Shah said. “The entire roundtrip is all happening in Voice Live API.” If a patient asks about information in their medical chart, Genie can also fetch data from their electronic health record (EHR) and provide answers.
Genie is a multi-tenant platform and allows each healthcare practice to customize its voice interactions with patients. For instance, the practice can upload information about the insurance it accepts, its hours of operations, and the types of procedures it offers. That information is stored in an Azure Cosmos DB vector database where the large language model (LLM) can retrieve it to answer common questions. In addition, Microsoft SQL Server stores all configurations and metadata across tenants. Azure Kubernetes Service serves as the underlying infrastructure layer for deploying AI services.
“We’re looking forward to using Voice Live API so that when a patient calls, Genie can detect the question and respond in a natural voice in near real time.”
Sidd Shah, Vice President of Strategy & Business Growth, healow
Keeping latency low was a big factor when the healow team was designing the voice capabilities of Genie.. “Because the Voice Live API is integrated into one single, unified solution, we can deliver speech-to-text and text-to-speech in the same infrastructure,” said Bhawna Batra, VP of Engineering at eClinicalWorks, which is partnering with healow to launch Genie’s voice capabilities. “If we did multiple hops to go across different infrastructures, that would add up to a diminished patient experience.”
In addition to integrating with Azure, Voice Live API in Azure AI Foundry stood out because of its support for multiple languages, noise suppression, and fine-tuning capabilities around pronunciation and voice. This gave the team more control over orchestration and how AI voices sound. It also offers multiple fully managed LLMs so healow had options to choose from, but didn’t have to deploy models, handle capacity planning, or provision throughputs.
As healthcare providers continue using Genie’s voice features, the healow team will measure its performance by tracking latency, as well as how many concurrent calls it supports and the average duration of those calls. “We want to consistently provide physicians and their patients an experience that is second to none,” Shah said.
Discover more about healow on Facebook, Instagram, and LinkedIn.
Follow Microsoft