Industry insights: How AI will revolutionize walk-in clinics

To kick off our Industry Insights story series, we talked to Arya Health’s Sam Gharbi for his perspective on how AI will revolutionize healthcare. Now we’re zooming in on a specific aspect of the industry: walk-in clinics. To do that, we caught up with Thomas Jankowski, the CEO of Medimap: an online search tool to help people in B.C. find an open walk-in clinic in their community.

The conversation has been condensed and edited lightly for clarity.

Vancouver Tech Journal: Can you describe what an average walk-in clinic’s relationship with technology is like?

TJ: To borrow an infamous line, “it’s complicated.” Doctors are schooled in medicine, not in the running of a modern-day business. Many doctors care about their clinic running adequately well, but too often this means implementing electronic medical records (EMR) they’ve been trained on or taken over from a predecessor, plus some form of a basic website, and nothing beyond that. I’d estimate that they trail the general small-to-medium-sized-business market by five to 10 years.

VTJ: Why is the technology so far behind?

TJ: There are many factors at play here. Running and operating a business is complicated, from building a modern, efficient, optimized website; to implementing a great online booking engine; to figuring out customer management, loyalty, and attrition optimization — the list just goes on. Some of these topics are actually more complicated than the day-to-day medicine they perform, and are constantly changing. At the same time, most doctors and their assistants don’t have the time to take a step back and evaluate the landscape. Finally, the primary software doctors use, an EMR, often has high switching costs or vendor lock-in. That means once they’ve set something up, they rarely have a chance to try something new, even if it could make their lives significantly easier.

VTJ: Can you describe what Medimap does?

TJ: First, Medimap is a marketplace (think Uber) that connects patients looking for non-emergency care with the right type of facility. We aggregate the largest number of walk-in clinics in the country, and as of recently, also the largest number of allied health practices (massage, chiro, physio, and more than 30 other verticals).  We also started adding pharmacies to our platform earlier this year, allowing them to capitalize on the recent changes where pharmacists can directly treat and prescribe for 14 minor ailments. What makes us amazingly convenient for patients is the fact that we’re the only place that aggregates appointment slots for allied health, and famously, wait-times for clinics.  From that point of view, we act more like a data aggregator (think Kayak). 

Secondly, Medimap is a super-directory (think Yelp). We optimize our clinics’ profiles so well that our listings usually rank much higher on Google, giving you access to those coveted clicks.  We also rank incredibly well for generic terms, like “physiotherapy near me”, which is how people tend to search most often.

Lastly, I’d like to think that we provide an unofficial public service, by load-balancing the entire healthcare system — and we do that by algorithmically prioritizing places with capacity over those that are too full. We refer to this as maximizing marginal healthcare capacity — and I think that if we were allowed and invited to actually do this at a provincial or national scale, we could drastically improve some of the problems within the system while saving billions of dollars as a result.

VTJ: How are you incorporating AI and machine learning into the business?

TJ: There are a number of ways in which we’re incorporating both. First, we’ve always tried to use a basic classifier or recommender system that allows you to see composite search results when you type in a symptom, in order to surface results you might be unaware of as a fit for your condition (such as a naturopath for some types of headaches). We’ve also recently launched a symptom-checker product that uses a massive machine learning dataset in order to provide a “light” diagnosis without leaving your home, or to give doctors a pre-triage tool to maximize on-site efficiencies. We also use a combination of different large language models like ChatGPT and Bard in order to help build out better profiles for our clinics to maximize their exposure. But we could be doing much more, because we collect a lot of data we haven’t done anything with yet. For example, in a recent analysis of some historical data, it became apparent that Medimap could’ve predicted COVID over a month before it became official, by looking at our internal search trends over time and seeing massive blips in cold-like symptoms versus historical patterns. 

VTJ: How have your clients felt about these integrations?

TJ: As a marketplace, we have two client bases: B2C — people seeking care — and B2B — companies offering healthcare-related products or services. Our B2C audiences seek to maximize convenience. Through our AI-based tech, we surface optimal appointment windows and wait-times, provide alternatives when it makes sense, allow you to check your own health, seek a different mode of care, and more. People absolutely love it and often can’t believe they haven’t heard about us before. Our B2B audiences seek to maximize exposure, minimize their overhead costs, and help with non-core tasks. We do that extremely well and always seek to raise the bar. As for extremely busy clinics that don’t need help with new clients, we help turn traffic away and manage their reputation because nothing can kill a business faster than an onslaught of one-star reviews.

VTJ: In many industries, AI is set to take away jobs through automation. How do you envision AI will affect the roles of walk-in clinic staff or doctors?

TJ: While concerns about AI are valid in some industries, it certainly doesn’t feel to be the case in healthcare, where constrained supply has been one of the main reasons the system is so broken and with demographics pushing this equation further south. We feel that AI, when properly deployed, can massively improve the myriad inefficiencies in the system, allowing us to squeeze more out of existing supply, lessen the burden on our suffering healthcare workforce, vastly improve patient communication and navigation, and yield significantly better health outcomes. All of these solutions are achievable now, even before we start to implement our blue-sky ideas: AI as a preventative-care tool, coupled with various IoT devices.  Without a doubt, AI will improve the quality of life and care for staff, doctors, and patients alike.

VTJ: How have investors received your tech?

TJ: Our investors are our lifeline — and never more so than in this day of vastly different economics than just two years ago. They’ve always seen Medimap as part of a master solution to a number of different problems intrinsic to our healthcare system. We’re already the front-door of choice to many people, who prefer to access Medimap than another government website, or would rather use our service than call multiple clinics to find what they need.  But our investors also see Medimap as the eventual one-stop shop for all things healthcare: a super-app that aggregates as much of the system as possible, and presents it in a way that maximizes convenience for the end-user while trying to intelligently source supply from every corner of the country.  And that’s exactly what we’ve been trying to build.  

Get in-the-know about innovations in your own backyard. Subscribe or become a member: