Contact centers – the perfect proving ground for AI in healthcare?

That's the message Talkdesk VP Patty Hayward is sending at HIMSS24, because, she says, contact centers have a high impact on patient experience, offer nonclinical use cases for AI, and keep humans in the loop.
By Bill Siwicki
01:21 PM

Patty Hayward, vice president and general manager for healthcare and life sciences at Talkdesk

Photo: Patty Hayward

Contact center vendor Talkdesk is placing a big bet on generative AI, transforming its technologies and processes with the AI that exploded in popularity with the release of ChatGPT.

Patty Hayward, vice president and general manager for healthcare and life sciences at Talkdesk (in Booth 1991 at HIMSS24), said genAI has a clear and nonclinical use case in contact centers – which is perfect for healthcare since experts agree it's best to prove genAI in nonclinical settings first.

She also said that contact centers are great for genAI because there naturally are going to be humans in the loop at all times. Experts agree one doesn't want to just throw jobs in the hands of AI without human beings monitoring and approving the work.

We interviewed Hayward for a deep dive into genAI and to find out what her main message is to attendees of HIMSS24 in the exhibit hall.

Q. Why did you turn to generative AI to bolster your technology?

A. Generative AI has had a massive impact on the customer experience market across all industries. Healthcare organizations we work with agreed, but to gain the benefits of AI in their contact centers, they needed a solution that accounts for their unique requirements and workflows.

So, we integrated genAI capabilities into our contact center platform, Talkdesk Healthcare Experience Cloud, via our Talkdesk Autopilot for Healthcare virtual agent – a connected, intelligent assistant that quickly and accurately resolves the most common questions and needs of patients and plan members.

Autopilot supports the entire patient/member journey through healthcare-specific integrations, workflows, and genAI models developed based on the company's extensive experience with healthcare organizations. Autopilot connects with electronic health records and other key systems to autonomously resolve complex needs and questions throughout the patient journey, from appointment management and symptom checking to revenue cycle and patient services.

We also use genAI to help our support agents do their jobs better and more efficiently. Let's say a caller simply wants to refill a prescription. Our contact center platform can instantly gather and organize EHR information about the caller, along with customer interaction records across channels and notes from those interactions.

AI is listening in as a copilot for the agent, pulling up recommendations and suggesting answers based on the organization's knowledge base. All this is happening while the agent is talking to the caller. GenAI empowers agents to become instant experts in the consumer they're serving and the specific questions they're handling.

Q. Generative AI may be hot, but it also is new. Do you think it's truly ready to handle this kind of work in healthcare?

A. No one is going to trust genAI outputs in a clinical setting today without some way to validate the data because these algorithms still struggle with "hallucinations" and incorrect information. You can't have that in the emergency department or operating room.

That's why it's absolutely critical to use genAI only in conjunction with humans in the loop. Contact centers are natural proving grounds for AI; they're critical to the organization, but they avoid some of the challenges around clinical decision making and have humans naturally looped into automated workflows.

GenAI is more than capable of enhancing customer support for payers and providers, who are beginning to use the technology in partnership with humans – rather than replacing them – to improve the patient/member experience and drive better outcomes.

We designed Talkdesk Autopilot to perform tasks patients request, but also to seamlessly bring in human agents when necessary. We make it easy for nontechnical staff to monitor and optimize how genAI works in their contact centers, training and augmenting the model as new opportunities or challenges arise with clicks, not code.

Q. What is your main message to HIMSS24 attendees on the exhibit hall floor?

A. As I mentioned, the contact center is the perfect proving ground for AI in healthcare. It has a high impact on patient experience and operations, there are clear and nonclinical use cases for AI, and there are natural human-in-the-loop processes. And it's an area in healthcare that is in desperate need of transformation.

But I want to be clear that our mission with AI in contact centers shouldn't just be to make things as fast and automated as possible. AI can absolutely create new efficiencies, and we do need them in healthcare contact centers. But we're talking about conversations that can be deeply personal, and some of them always require human interaction.

AI that can resolve transactional, high-volume chats and calls frees up human staff to be better in the conversations where they are most needed. And AI can take the massive amount of data that a provider or payer knows about a consumer and make it summarizable and actionable for human staff in real time.

That means they can be proactive, accurate and informed representatives of an organization responsible for that person's care.

When we think about a near future where AI is filling that role – instead of just deflecting as many callers away from human staff as possible or focusing on making agent conversations as fast as we can – the position of the contact center will shift. It won't be seen as a cost center, but a real driver of growth and better outcomes for patients and members.

Follow Bill's HIT coverage on LinkedIn: Bill Siwicki
Email him: bsiwicki@himss.org
Healthcare IT News is a HIMSS Media publication.

Want to get more stories like this one? Get daily news updates from Healthcare IT News.
Your subscription has been saved.
Something went wrong. Please try again.