After raising $3 million in seed funding led by Canvas Ventures, startup Parakeet Health launched a generative AI voice platform focused on patient engagement. Two of the San Francisco-based company’s executives recently spoke with Healthcare innovation about what it means to bring generative AI to doctor’s office call centers.
CEO Jung Park previously held executive positions at One Medical and Epocrates, and co-founder Eric Mao led engineering and product teams at Microsoft and Primer AI.
Healthcare Innovation: Why did you decide to focus your generative AI efforts on the patient-facing aspect of healthcare?
Park: People may think that focusing on using AI to solve management challenges is not so attractive, right? But if you think about it, implementing AI clinically has its own challenges. One of the things I learned when working on Epocrates was not to change the workflow. So if you really try to do something that replaces what a doctor does, you may not necessarily get the adoption you’re hoping for.
HCI: What are some of the challenges physician groups face when managing patient contacts and calls?
Park: It’s something I think we can all relate to. It is not unusual to wait five minutes before answering a call. Nobody wants to be left waiting. Everyone’s time is valuable, and if these issues are not addressed, patients feel like they are not valued. Additionally, for many medical offices, staffing is the biggest challenge. There is rotation. Labor costs continue to rise, and for people who actually address many of the support issues, the concept of empathy fatigue is real. Very often you can have a smile on your face and do it consistently, day after day. At some point people get exhausted. That’s a problem we all face as healthcare administrators.
Another challenge, I would say, is revenue leakage. When running a medical organization, you have many opportunities to generate income, but there are also many opportunities to lose income. Let me give you an example: When you have a late cancellation, that means the patient who is supposed to come and take that provider’s schedule doesn’t show up. So it creates these gaps in the schedule, which results in lower income.
When a specialty practice receives a referral from a primary care practice, they need to convert this referral, because they are like a golden ticket, right? If a doctor tells you: I would like you to go see this dermatologist about this rash, then you go. But it turns out that people don’t always go. So when you can actually convert all of these referrals, that’s really the way you’re going to generate income.
HCI: So how is Parakeet working to address these challenges?
Park: First, it improves the patient experience by answering every call 24/7 with no wait time. Parakeet also automates repetitive tasks like scheduling and answering frequently asked questions, so that staff, instead of doing these repetitive tasks, can focus on more valuable things that patients need, like checking in, checking out, and actually addressing patient needs. when they are actually in the physical office.
From a revenue leakage perspective, Parakeet actually makes outbound calls. So if someone cancels, Parakeet can call the next person on the waiting list, or maybe there’s a patient two weeks from now who might be available to come in and fill that gap in that schedule. That’s one way to bolster provider utilization and ensure that revenue is captured for the healthcare organization.
HCI: Over the years, most people have probably had experience with a chatbot on a retail or airline website and are often frustrated because they can’t answer their question. But has technology improved exponentially now?
mao: I’ve been in AI for a long time. I don’t think it’s ready to replace doctors, but we see firsthand that it can now address these challenging programming questions, and I think it’s ready for prime time for this set of use cases. Transcription quality is now much better with this wave of great language models. And the speech sounds like that of a human, which is a very recent thing. It’s in the last two years that we’ve had this.
HCI: Does your system have to integrate with the practice’s EHR, billing system, and customer relationship management system for this to work?
mao: We spent the last year building all this infrastructure to achieve this. We integrate with your EHR, with your phone system, with your CRM ticketing system, all of those things to enable that very natural experience. So when you request an appointment, it searches for it in real time, gives you real-time availabilities, and then we book it directly in the system so that no human interaction is needed throughout the entire process.
HCI: Is that process a challenge or do all of these systems now have APIs that make it work seamlessly?
mao: Hopefully so. But we’ve found that a lot of these EHRs don’t have very good APIs, probably sometimes on purpose, because they want you to do everything on their platform. But we put a lot of engineering effort into making it reusable. We’ve made it easy to onboard the next client, because we’ve already done all the heavy lifting.
HCI: It sounds like some of the clients you’ve announced are virtual. Have you initially targeted those types of companies?
mao: We go to doctors’ offices that usually have a call center. So we don’t address mom and dad. When you have a call center with 10 or more people, you will see more cost savings and more use of what our platform does. We have dermatology; We are in conversations with OB/GYN and primary care.
HCI: So, would it make sense for a large health system like Sutter Health or Kaiser Permanente?
mao: We will focus on the Sutters later. We’re targeting outpatients first, but that’s absolutely on our roadmap to address Sutters, UCSF and university health systems.
Park: We are already seeing that our solution has hit a sensitive spot. Whether it’s those people who actually work more on the traditional side or in digital companies, at the end of the day, the opportunity to interact with patients is a real thing. Ninety percent of us still use the phone to make an appointment, which means there is an opportunity across the board to actually interact with patients.
HCI: Are traditional vendors in the patient engagement space also trying to bring AI solutions to market?
mao: There are incumbents in the patient engagement space, but they are using old technology. They are trying to direct this giant cruise ship to use LLM, while we are native to LLM. We started day one with this technology, so our technology works better. Over time, incumbents will try to adopt this new technology, but we believe we can move faster and our technology is already here using those things.
Furthermore, the competitor is offshore call centers. When we talk to clients, when they have problems with too many missed calls, they look to expand more people through offshoring to solve that challenge. They’re not even looking at the technology. So when we show them what we have, they say, ‘Wow, I didn’t even know this was possible today.’
