Leveraging AI to Increase Clinical Capacity by 500% with Muthu Alagappan of Counsel

This is part of our executive insights series where Elion CEO Bobby Guelich speaks with healthcare leaders about their tech priorities and learnings. For more, become a member and sign up for our email list.

Role: Founder & CEO 

Organization: Counsel

What led you to launch Counsel?

We founded Counsel because we were excited by the opportunity for AI to move beyond automating administrative work to supercharging clinical interactions. When you think about solving a supply-demand mismatch in healthcare, the most robust way to do that is to multiply the supply—multiply the world's clinical capacity. Too many people are working on marginal improvements; but instead of improving metrics by 10-20%, we’re exploring how to make doctors 500% faster or 1,000%. 

The bet we’re taking is that by introducing AI clinicians into patient care in a way that's safe and maintains a doctor in the loop, then that is the best of both worlds. We’re taking asynchronous care and sort of introducing AI in a very meaningful way to allow patients to really have unlimited access to doctors. 

How do you think about what can be covered by this asynchronous care model?

There's a few things that we've been really excited about:

• If telehealth was about doing in-person care virtually, then the first iterations of Gen-1 asynchronous care were really about taking urgent care and moving it from video to text.

• Beyond that, we’re broadening the use cases for asynchronous care. For example, a lot of the questions we get are actually more advice-focused, things you wouldn't necessarily go to Teladoc or urgent care for; it’s often questions people currently go to Google for. 

• There are many vertical, specialty-specific point solutions, and we think what’s been missing is a core orchestration layer, similar to a PCP, that captures patient interest at a moment of need and helps route them to the appropriate care in a really consumer-friendly way.

The other differentiator for Counsel is that we have an AI-first perspective. We do have our own in-house medical group, but by the end of this year, about 90% of the clinical interactions will be primarily done by AI. It's really the evolution of asynchronous care—care performed by AI and augmented by doctors, rather than performed by doctors and augmented by AI.

Is it too strong a framing to say you see this as a PCP replacement?

I’m an internist, and I can't tell you the number of clinical questions we get, that if asked in my clinic, I would have either no idea what the right answer was, or very little grasp of the knowledge-base required to answer this question. The amount of medical information is growing rapidly, and no one can keep up. But because of our care model, we're giving best-in-class answers based on the latest evidence at the point of need within minutes, which PCPs simply can’t do.

But beyond that, I think the PCP model as it is today will not survive. I don't know what the time frame is, but it will probably be disrupted and replaced by something that is an order of magnitude better. You can imagine telling our children in 20 years, “We used to have this concept of primary care physicians. It was a single person. They could only see 10 patients a day. You usually couldn't get in to see them when you had an issue. There weren't enough of them, and they had to try to remember everything about your health.” When that happens, we'd like to be one of the institutions that's proposing a better model.

I'm curious if you have a vision of how you see that shift happening over time, given the breadth of things that PCPs do today?

If you were to deconstruct a PCP’s role into, let's say, 20 different functions, we will likely see specific sub-functions be performed so much better in an asynchronous or virtual context that patients can no longer ignore it. It’s analogous to what’s already happened with hospitalists taking over inpatient rounding in the 90s and virtual care taking over certain urgent care functions. We think PCPs will continue to give up discrete functions as point solutions deliver 10x‑better experiences. 

How do you talk about “AI replacing doctors” without alienating clinicians or regulators?

First of all, we anchor ourselves as both an AI company and a medical practice. Patients care about outcomes and experience, not the mechanism. Internally, we keep physicians deeply in the loop—both to ensure safety and to build trust. Culturally, the industry is moving faster than many realize; at an event recently, the idea that AI clinicians will be mainstream within a decade felt mainstream, not fringe. 

I think looking at utilization patterns of younger generations is a good way to predict where the world is headed. If you look at a lot of people in their 20s, they probably do go to ChatGPT as their first point of contact for medical questions. So whether or not we philosophically think AI can do doctoring, that's how people are already starting to use it in generations that will predict future behavior.