Joey Wahler (Host): We're discussing using technology to solve healthcare operational challenges. Our guest, Mohan Giridharadas, he's founder and CEO of LeanTaas. They're a healthcare analytics company using artificial intelligence with 350 employees based in Silicon Valley in Charlotte, North Carolina. LeanTaas is also one of ACHE's premier corporate partners who support ACHE's vision to enhance healthcare leadership excellence.

This is the Healthcare Executive Podcast, providing you with insightful commentary and developments in the world of healthcare leadership. To learn more, visit ache.org. Thanks for joining us. I'm Joey Wahler. Hi there, Mohan. Welcome.

Mohan Giridharadas: Hi, Joey. It's nice to be here.

Host: Great to have you aboard. So first, you got an electrical engineering degree in your native India before earning a MS in Computer Science from Georgia Tech, MBA from Stanford. So, when did applying that great background to improving healthcare systems first pique your interest?

Mohan Giridharadas: It was a long journey, Joey, as you indicated. I spent several years as a software engineer at Intel and at Mentor Graphics before deciding to go to business school at Stanford. When I got my MBA from Stanford, I joined McKinsey and the original plan had been I would be at McKinsey for three or four years and then start a software company. It turns out I enjoyed McKinsey and the work a lot. So, I ended up staying 18 years and, in the process, was running McKinsey's Lean Manufacturing and Lean Service Operations practices, both in the U.S. and across all of Asia Pacific, which I did from Australia and Singapore.

I left McKinsey at the end of '09 to start an analytics company, because here was the realization I had, that whenever I saw operational excellence initiatives being done, they were being done on the backs of Excel spreadsheets. And in my mind, Excel was middle school math. And if you use middle school math, you'll get middle school results. And so, my idea of LeanTaas was imagine if you could deliver operational excellence, but do it replacing simple math with sophisticated math, optimization, simulation, AI, machine learning, et cetera, and build it on a software as a service platform. So, that's why I named the company LeanTaas, for Lean Transformation as a Service.

Initially, we were across many industries. We had Home Depot, Flextronics, Google, Clorox, PNC Bank, et cetera, as clients. And what we were doing, the common element was we were absorbing enormous amounts of data on a continuous basis. We were building sophisticated algorithms to improve operations. And we were delivering the results over a web-based application that users could engage with. And so, that was the business model.

And then, we had a bit of a seminal moment in July of 2013. I was sitting in the office of the guy who ran Cancer Services at Stanford HealthCare, and he still runs Cancer Services at Stanford HealthCare. He's an ex-GE guy and understands lean and process flow and so on. And he posed a very simple question. He said, "You guys know lean, you guys know AI, machine learning, and software. Answer me this, why does my infusion center look like a ghost town early morning and late afternoon, but is like a train station in the middle of the day where all the chairs are full, the waiting room is bursting, the nurses are scrambling? And yet we can't seem to get the flows better." That seemed like an interesting problem to work on.

And so, we worked on it together. And it took us nearly a year to get it figured out. And another six months or so to get it running properly. But we had magic. The wait times went down by 50% for patients. The ability to serve more patients went up by 15% or 20%. And nurses stopped missing their lunch break. So, we knew we had magic in a bottle here and said, "We should get out of the other lines of business and focus the entire company on delivering this lean transformation as a service into healthcare, and that's it." So, we pivoted to entirely infusion. Those days, it was one infusion center, 50 chairs. Today, 700 infusion centers, 14,000 chairs, 30-35% of the U.S. capacity in chemotherapy happens on our platform. We then expanded and added additional products like the OR product and the inpatient product. So, that's the trajectory of how we got here.

Host: Wow. So, that's quite a story. Why would you say, Mohan, it's so key that AI be involved in solving healthcare operational challenges rather than relying on current workflows or systems like the electronic health record or EHR as it's called, that's been traditionally used?

Mohan Giridharadas: So, couple of things, one, if I think about healthcare in the U.S. over the last 30 years, the clinical advancements have been miraculous, right? Robotic surgery, genomics, precision medicine, proton beam therapy, these are all magical things that no one could have imagined could be built, much less deployed. Meanwhile, the operational sophistication in healthcare has not really advanced. Forty or 50 years ago, it was paper-based records, and yes, those have been digitized. But think what they've been digitized on. They've been digitized on EHRs, built by companies that were founded 45 to 50 years ago. That is the reality of it.

And the mental model for how EHRs think about operations is just mathematically incorrect. Everything in the EHRs is a resource, and you reserve a resource. That's a deterministic way of doing things. That's the way you book a tennis court. So, Joey, if you have the tennis court 8:00 to 9:00, and I have it 9:00 to 10:00, we both understand I'm kicking you off the court at 9:01. Healthcare doesn't work like that. Nothing starts on time, nothing ends on time. So, you need sophisticated math to help probabilities, cancellations, flex times, and so on to run the optimization. And so, nothing in healthcare's operations were built with the understanding of how complex life would become, right?

So, think about Delta Airlines. Delta Airlines in the '30s had two flights a day from Atlanta, one going west and one going east. So if you ran airport operations in the '30s, you would say, "Joey, take gate one. Bob, take gate two," or "Joey, your flight's late. Go help Bob on gate two," and that was airport operations. Today, Atlanta Airport has 3,000 or 4,000 flights a day. Delta has a thousand flights a day out of Atlanta, 200 gates. The "Joey go to gate one, Bob go to gate two" doesn't work. You need algorithms to balance the supply and demand math, to cover both sides of the equation, to understand network topologies and how patients flow through the various services. It takes crazy math to do this. And that is completely non-existent today, and that's what we are changing.

Host: Now, you have a flagship suite of products that enable health systems to improve patient access and lower costs by unlocking capacity and what are known as scarce assets. So, what is that and what does that mean for your clients?

Mohan Giridharadas: So, we started by focusing on the assets that are valuable and are difficult to schedule. So, infusion chairs are one example, operating rooms are another example, they're the financial lifeblood of a hospital, and inpatient beds are an example. So, here's the reality of what happens. When capacity gets squeezed, everything comes to a halt. So, the simplest way to think about this is the freeways. At midnight, the freeways are not a problem. You can go from point A to point B very, very quickly with no problems. But during rush hour, the capacity of the freeway has reached the maximum. There's not enough concrete to put another car on the road. Therefore, it's complete gridlock. Everything gets a hundred times worse. A lane change takes a long time. A fender bender would delay a hundred thousand people for hours.

And so, hospitals, unfortunately, are operating at the edge of capacity. And the sea change that has occurred is 30 years ago, it was easy to just build more. Build more hospitals, build more ORs, build more ambulatory clinics, buy more imaging machines. That reality has changed. The hospital's financial margins are under pressure and they can't just build, build, build anymore. So, they need to do more with what they've got. And to do more with what you've got, you need to have the sophistication of what we provide. And so, that's why we are focused asset by asset to improve the performance.

Host: And so, your products now, Mohan, have been deployed, I understand, at more than 800 hospitals and centers and 100 leading health systems across the country. How would you sum up what their response has been to all this?

Mohan Giridharadas: Enormously good. We have very high loyalty with our customers. They've stayed with us forever. By the way, we don't just say it, we prove it. All of our contracts are written with 100% money-back guarantee, meaning after we get started, for any time within the first six months, if they say this is not working for me, either because I don't need it or my people can do it or it isn't working as advertised, we will refund every nickel they've paid us. That is an enormous step that I would argue virtually no other healthcare vendor does.

Second, we don't confuse customers with hostages, meaning if they don't like our product, they should have the right to cancel anytime, as opposed to being a hostage for the next five years of the contract. So, all our contracts are written with a cancel-anytime clause. In the beginning, we had to do it to give hospitals the reassurance that we would be there for them. But long after that, we continue to do it because we fundamentally believe that we have to earn the right to serve our healthcare customers every single day. So every day, our product has to work, our teams have to work, we have to be responsive, and we have to be continuously innovating. If we don't, they have the right to cancel at any point in time. So today, 14,000 infusion chairs woke up, 6,000 ORs woke up, 25,000 beds woke up, and you know what? Nobody canceled. And the same thing will happen tomorrow.

And if you think back to the pandemic, the health systems were under enormous pressure to save money. It was literally burn-the-furniture-to-stay-warm kinds of operations, right? We were the only vendor that had the cancel-anytime clause. So, we had the bullseye tattooed on our forehead as the easiest vendor to get your money back from or to cancel your contracts with. And yet, only a few health systems cancelled during the pandemic because they were under enormous pressure. But by and large, they all stayed with us.

The good part about doing operational things is the results are there in front of them. They can see the better utilization. They can see the shortened lead times. They can see the ability to get an OR block whenever you want. They can see that they're doing more discharges a day, and that the discharges are happening earlier. So, the metrics are very easy for them to see. And that's what keeps our customer loyalty as high as it is.

Host: Well, that's certainly standing behind your product. And you mentioned being innovative there, Mohan. Speaking of which, do you think AI will live up to the exciting potential that many feel it has in the near future? What are you hearing from the health systems that help shape your work as you move forward?

Mohan Giridharadas: AI is here to stay, but it'll take time. Everyone wants it right here, right now, but it'll take time. And the way we think about it is there's clinical AI. And then, there is operational AI, right? So, clinical use cases, in our mind, are still a ways off. It may start in narrow areas like reading scans or reading patient notes and creating summaries and those sorts of things, and maybe on the administrative burden side by providing ambient scribes, AI assistance, those sorts of things. That's kind of where we will see AI enter the mainstream.

We are focused on operational AI. So, the way we think about it is imagine having an always available operations expert embedded into our products. So, we call it iQueue Autopilot because our flagship platform is iQueue. iQueue for infusion centers, iQueue for operating rooms, iQueue for inpatient flow. And iQueue Autopilot is our ever-present generative AI expert. And so, all our applications will have an Ask iQueue button where you can press the button and either Lean Forward and ask it questions. "Dr. Smith wants a new OR block. Should I give it to him? Yes or no? Who should I take it from?" These are difficult questions, but they are analytically answerable, and asking our autopilot will get you the answer to that. That's the Lean Forward mode.

The Lean Back mode is you didn't ask me a question, but I think you need to know the answer to this, meaning, let's say you're a nurse manager controlling unit two, and we can see that eight days from now, you're going to have a shortage of nurses on your unit, it's good we tell you now, so you can do something about it. So, that we call Lean Back. You didn't ask me a question, but I'm going to tell you the answer because I think you need to know the answer. We think of that like your security system at home. You didn't wake up at 3:00 in the morning and say, "Is my back door secure?" But if for some reason somebody kicked down your back door at 3:00 in the morning, you want your security system to tell you, "Hey, heads up, your back door's been kicked down," right? And so, we think of that as lean back.

And the third mode we have is in-the-moment problem-solving. You're stuck and you want to solve it. What should I do with this patient? Should I divert them to another hospital? Should I put them in this unit? Should I put them in that unit? Those are hard problems to solve. But if you had a button that says, "Give me some suggestions as to what I can do," and the AI could give you intelligent suggestions, that's fantastic. We think of that like your navigation system in your car. You're no longer worried about getting lost in the middle of the night in a part of town you're unfamiliar with because you can always hit go home on your navigation system and it knows to take you home. And so, those are the three modes in which we expect our Autopilot to be embedded in our products. By the way, we've already launched them, and it's starting to find its way across our customers.

Host: Switching gears a little bit, Mohan, we've talked obviously a lot here about technology, but along with that, what about the need to change management in transforming hospital operations to go along with that technology?

Mohan Giridharadas: That is an incredibly important thing. Change management is a significant requirement. You cannot throw software over the wall at anybody, but you really cannot throw software over the wall into healthcare because the staff is running around. They're not sitting at a desk gazing at a screen. They are with patients, doing patient care, making complex life or death decisions in some cases. And so, change management is a thing. People are used to the current way of doing things. And so, anytime you want them to do something slightly differently, it is a big deal. None of us can even brush our teeth with the other hand, if you think about it, right? That's as simple a change as it gets. But if I asked you to start brushing your teeth with your left hand, or in my case, I'm a lefty, start brushing my teeth with my right hand, I'll struggle with it. So, change management is a big thing.

But here's the problem. If you do what you've always done, you'll get what you've always got. And if you're not happy with what you're getting, you need to do something different. Otherwise, you'll keep getting the same thing. And so, what we do is we elevate the workflow in very important ways. One, we tell people part of the reason you're struggling is you're forced to react and you're playing whack-a-mole on the front lines of a hospital. Imagine if we could predict with great certainty what's going to happen. Think about this now. We look at our phone and we know it's going to rain this afternoon. And so, we know to take an umbrella or to stay indoors, which 50 years ago, nobody knew. They would just wing it and guess. So, we can predict with high precision what's going to happen. Instead of just reacting in the moment from gut-level instincts or looking at spreadsheets or downloading some report from the EHR, we prescribe what is the best answer. And then, we automate what needs to be done so that the staff can work at the top of their license and not do stuff like pulling numbers and sending messages. So if we do those three things, here's what happens. It becomes easier for them to do their job. It becomes more intuitive. They get better results, and they enjoy the experience more. And it reduces their stress and their burnout. So once they see that and experience that, the change management burden automatically drops away. Think about it.

In the beginning, I was used to dialing 777 to get a yellow cab. Then, there was a bit of a change management bump and I started to realize Uber can work quite nicely. Today, nobody from Uber is sitting in my house trying to change manage me into using Uber. It's just the best way for me to get a ride to the airport. So, we've understood this. The Taas in LeanTaas stands for Transformation as a Service. We've had that in our name from day one. And we've understood that in order for our products to have real impact, we need to be there in the trenches with our customers, helping them achieve that impact.

And so, the way we do it is we assign a dedicated team to every client for every product, who are deeply expert, and they stay with the client for the duration of the relationship. I told you Stanford got started with us in 2013, 2014, that was 10 years ago. We still talk to Stanford on a regular basis. And we still understand how the product can get better, what they can do. As they have turnover, we train their new people. That's all just part and parcel of our offering, and we believe it's so incredibly important, and it's such an intrinsic part of our service that we will never stray from that mission.

Host: A few other things. AI, as you know better than anyone, is such a hot topic right now. How would you say health systems should evaluate potential AI solutions amid all the discussion and debate out there?

Mohan Giridharadas: You're right. There is a lot of hype and there's a lot of noise and that creates AI washing, right? Everyone finds some reason to say they're an AI company, whether it's changing their name to be a .ai or doing something. So, you do have to, if you're a health system, not get caught up in the hype. Health systems are rightfully very conservative, and I think that's a great thing. If I'm undergoing surgery, I'm actually quite relieved that the surgeon isn't using a scalpel he found last night on Amazon. I'm happy that they are very conservative.

And so, my advice to health systems would be, one, ensure that the team or the company you're working with has real expertise, meaning they've got people with PhDs in mathematics and data science and so on like we do, and the real credentials to do this. This is not someone who discovered AI last week. You've got to then make sure that they have expertise in the use cases that they're talking about. That this is not a health problem, the healthcare problem that they discovered yesterday and they're starting to work on it. That they've got a depth of expertise that it takes.

For example, on infusion chair scheduling, we've been at it for 10 years. We've spent $200 million on the problem. We've got a 100% team that has done nothing but infusion schedule optimization for 10 years. The same with our ward product, same with our inpatient product. These things take time, they take money, they take people, they take effort. And so, you have to make sure that people you're partnering with have the expertise in the specific use case. So, just like we are not experts on everything, we're experts on the few things we do, you've got to make sure that that's there. And then, you've got to make sure there's a track record. You want to make sure that there's a proven story. Not promises on PowerPoint with excuses when it doesn't work. A proven story. Be able to tell them, show me 20 cases where this has worked. Show me those 20 cases where it has worked for one year, two years, three years. Not a johnny-come-lately pilot. Remember, the first canaries in the coal mine always die, so you want to make sure that they can point to a track record that works. And then, that they've got the infrastructure to make this and sustain this, meaning the data requirements, the data security, compliance, PHI, the model bias, all of that has to be there. How does the model learn? How does the model update? Is it a black box? Can it be explained? So, those are the kinds of things that I would encourage health systems to think about.

Host: And then finally, in summary here, Mohan, you also co-authored a book on hospital operations called Better Healthcare Through Math. So over the next year or so, what do you think will be the single biggest challenge hospital operations teams will face and how can those leaders stay ahead of all that?

Mohan Giridharadas: Single greatest challenge. Would you let me have two challenges?

Host: Sure.

Mohan Giridharadas: Okay. So, I think one big challenge is going to be staffing. Staffing has been a problem and will continue to be a problem. And it just makes it harder to capture the full utilization potential of the assets, right? If you cannot staff an OR, you cannot do surgeries in it. And so, hospitals will have to continue to innovate in this area. Hire more, train more, automate more, use GenAI, use better tools to predict and assign staff, et cetera, et cetera. So, just getting the staffing nut cracked because the shortages across all skills, nursing, physicians, et cetera, is bad and projected to get worse. So, that's challenge number one.

Challenge number two, I think the financial pressures on the operating margins are going to cause hospitals to tighten up. And when they tighten up, they'll focus on cost reduction, not value creation. And the problem with that is it leads often to a penny-wise mindset. So for example, EHRs always have claimed that they can solve every problem under the sun in healthcare, right? And the moment the EHRs do that, the hospitals freeze because they say, "Hey, I've spent a lot of money on the EHR. So, I should get the most out of it." Makes all the sense in the world. And therefore, they don't want to spend money on third party solutions, and so they wait, and they wait, and they wait. And therefore, they don't capture the art of what's possible. And the reason this makes no sense is it's a bit like saying, "Hey, I spent $50,000 on a car. I don't want to spend any more money on a boat. So if I want to go water skiing, I'm going to wait until the next upgrade of the car allows me to water ski with it." you're going to be waiting a long time, right?

So, what I believe hospitals need to do is to recognize we are at an inflection point where the opportunity to capture significant advancements in operational excellence is here right now. And waiting is an opportunity cost. In some ways, it's like the inflection point at the start of the internet, right? I'm old enough to remember when there wasn't an internet and now obviously when there is one. Imagine back then if we had said, you know, IBM's got it figured out. We're going to just wait until IBM is going to deliver the entire promise of the internet to us. Imagine if we had all said that. No Google, no Amazon, no Facebook, no Instagram, no Netflix, no Uber. That's not the internet we know. And so, this is essentially what's happening when health systems blindly believe the EHR vendor is going to take them to the promised land. It's not going to happen. Every solution takes hundreds of millions of dollars. The transfer to the promised land will be done by 10,000 software companies, 9,500 of whom haven't been born yet. And that's what it's going to take. The EHR vendors have hundreds of things to build. They cannot invest the level of time and effort it takes on every single thing to build and deliver the value. And so, in the meantime, hospitals will wait and they will suffer and the patients will suffer. And so, that's the thing I would encourage hospitals to up their game in terms of how they think about it.

Host: Well, folks, we trust you're now more familiar with technology to solve healthcare challenges. He calls it magic in a bottle and he's pouring it very successfully all over the healthcare industry, Mohan Giridharadas. Exciting stuff. Keep up this great work. A pleasure. Thanks so much again.

Mohan Giridharadas: Thank you for having me, Joey.

Host: And for more information, please visit healthcareexecutive.org. If you found this podcast helpful, please do share it on your social media. I'm Joey Wahler, and thanks again for being part of the Healthcare Executive Podcast, providing you with insightful commentary and developments in the world of healthcare leadership.