Bryan Schielke 0:00

So, John, you're building digital twins of populations to guide real world decisions. When did you realize simulation was not just an academic tool, but a leadership advantage?

John Cordier 0:08

Great question. As we were looking at the commercial applications of the software, understanding that there's some decisions that are meant to kind of be done by data scientists, but ultimately it's about improving the health of populations of people. And those decisions are made in the C-suite about what actually is going to get funded, where time and energy of an organization are going to go. So is this realization that movement from here's an academic tool to build understanding of a system or understanding of the world as it exists to now this needs to get applied and let's apply it in ways that can be positive for, you know, it's going to be growing revenue, growing market share of products and services that serve the health and well-being of a population of people. So um another way that that came up for us, you know, we had one of the best forecasting tools during COVID. You know, this was the work that went into this started in 2004. And so from 2004 until 2020, there was a lot of work that went into this. And even though we had the best forecasting system, didn't mean we're getting contracts, didn't mean that people were listening to the results of those forecasts. Who were they listening to? They were listening to people that had the ear of the C-suite, the Deloitte's, the McKinseys, the BCGs of the world, all known, trusted entities. But you might have the right tool, but if you don't have the ear of the right people, which you know, that's the kind of leadership advantage that you you mentioned inherent in the question. If you don't have both, you're at a disadvantage.

Bryan Schielke 1:40

Absolutely. That makes sense. All right. Uh next question here. Epistemic sits at the intersection of public health, economics, and AI. How do you help executives trust model-driven insights when the stakes involve lives, markets, and reputations?

John Cordier 1:55

Yeah, uh really good question. So the approach that we use is called agent-based simulation. And the difference between agent-based simulation and a large language model or a black box machine learning model or neural net is that every single interaction is transparent within an agent-based model. And what I mean by that is you can see here's the mechanism of how the system is put together. And when you overlay that system kind of model on top of the population, you can see are the people behaving in the way that are reflected in reality. And what's cool about that is you're able to understand the counts, the timing, and the interaction of influence on behavior at the population level. And so, you know, when people kind of talk about the intersection of public health economics and AI, like we're providing the population at a baseline, but then the agent-based simulation technique provides the transparency that you don't get with some of the other techniques. So we believe that's a real advantage, especially when sort of trust and credibility in data that you're using is being put into question by some of these other systems.

Bryan Schielke 3:02

Wow, that's amazing. All right. Thanks, Shandra.

John Cordier 3:05

Appreciate that. It's a really big deal for the trust from leadership as well. So, and it's both trust and leadership, and it gives the data science folks that are using these tools to drive better understanding about how the world works, it's giving them a voice at the C level as well. And so the ultimate goal of this system is we have AI tooling that enables more everyday people to build models that can then get kind of certified or verified by very astute data scientists. And by having this system in place where you have trusted models on the population, you're actually seeing the results come out in reality. Everyday people can participate in the decisions that are happening to them at state local levels, which is ideal, but also within organizations where you kind of enable the best ideas to go to the top. And this is a system that can enable that to happen.

Bryan Schielke 3:56

Well, that's huge. Fantastic. All right, good stuff. Okay. Next question. Your mission is to make agent-based modeling accessible to everyday decision makers. What's the biggest misconception leaders have about data-driven strategy today?

John Cordier 4:10

So the other key part of our mission is to increase clarity that advances health and economic outcomes. And honing in on the word clarity, I think is really important for any executive decision maker when they're looking at using data to help drive decisions. As a company, we define clarity as having the information to execute with conviction. Executing with conviction is a state of being. And if everyone's executing with a high level of conviction that's rooted in clarity, we believe that that gives organizations a competitive advantage. And even for ourselves, we evaluate how clear are we in the actions that we're taking on internally. So when thinking about data that drives clarity, we believe it comes in three forms. The information includes the intention. So like, is there a metric? Is there a set of outcomes that you're really driving towards? The second piece is context. So that's using data to have situational awareness of what are the conditions that you're operating in, and also having the context of if we make a certain set of decisions in the context of the conditions, what are the likely results going to be? And then the third part is what's the expected impact? So being able to, as a leader decision maker, play out these scenarios, it should give you more clarity as to here's the trajectory we're on. We make a decision in one way or another, here's how that's likely going to play out. And by doing so, some people say, like, oh, well, like the result is getting better forecast. It's like that's part of it. It's really trying to get clarity into what decisions are leading to better outcomes. And instead of focusing on here's the outcome that we're going to fixate on by getting at this certain point in time, it's really how can you make decisions that set up conditions that whether or not, you know, one right or wrong decision comes up, the conditions are moving you in the direction that lead you to where you want to go. So that's the kind of benefit of using this type of a system. And when thinking about data and decision making, how we kind of take a different approach to it. Oh, that's fantastic. Awesome.

Bryan Schielke 6:08

Good stuff. Okay. The next question. You started in public health and even founded a nonprofit focused on reducing disparities. How has that mission shaped the way you build a for-profit technology company?

John Cordier 6:20

I think like the underlying mission is pretty similar, like for me personally. There's a book that I read when I was figuring out what I wanted to be when I grew up called The Health Gap. Um, it's written by this guy, Michael Marmot. And what he talked about is the conditions that people live in usually lead to the outcomes that we see in health or economic mobility or whatever else. So I think thematically in my life, I've been choosing things that work to address improving conditions that give people more agency. And so the nonprofit that I had set up, it was selecting a health disparity, you know, enabling high school students to select a health disparity in their local community, do a research project on it, understand it. But what we were really teaching is you can make an impact in your community. And like the leadership skills to do that is what we were really driving towards. Health was one way of connecting with people because we all have some person in our life that we can think of and it's like, oh, you've dealt with some traumatic health issue, or you know, you you know you have somebody in your class or you know, a family member, whatever it might be. And so I think when I look at what we're doing with epistemics, we're really trying to get a technology into the hands of more people so that they can understand how to improve health and economic outcomes at a local level for the populations that they care about. So the same mission underpins both of those, I think. And just like the means of delivering on that mission, like from a technology company. You know, we weren't in a position to like get funding unless we were showing okay, people like want to adopt this, companies are able to benefit from it. That's all fine and good. Ultimately, we're trying to get this technology into the hands of the local government folks, state government folks, NGOs, nonprofits who actually want to like put their do good energy into action and see results within their lifetime. That's the kind of stuff that I think everyone who works at Epistemics gets pretty jazzed about enabling. That's awesome. Yeah.

Bryan Schielke 8:21

Complete win-win. I love that.

John Cordier 8:23

Sometimes there's a trade-off between focusing on the health, the economic incentives, or a political incentive. And our objective is to enable health and the economic things to play out in a better way. Because like if you're able to improve the economic mobility of people, that increases their agency and that'll also have a positive impact on their own health eventually. And so we believe we can at least have if you focus on like those two things, good stuff will happen. That's amazing.

Bryan Schielke 8:49

Very cool. All right. The next question here. In moments like pandemics or market disruptions, leaders often default to instinct. How do you coach them to balance intuition with simulation without losing speed?

John Cordier 9:02

Great question. We haven't totally figured out the answer to this just yet. But the challenge, again, comes from like when that change happens, data from the past becomes a less reliable predictor about what's going to happen in the future. And so if you can understand the behavior that's changed, ideally you can start seeing, all right, here's what might happen or what might play out. And that's the key part of using a population model where you can like run these simulations to test counterfactuals. So it's like, oh, we don't know exactly what's going to happen. And you look at whatever is in the news cycle every other week or every month. There's like, here's this new crisis that the world is going to be dealing with. And how are people thinking about it? How are people behaving differently because of it? Those are the things that you'd want to start running simulations of to see what trajectory we're actually going to put ourselves on and how we respond. So people on it, it's really just like giving them the tool ultimately. So like making it accessible rather than, you know, it's just a tool that data scientists are using, making it accessible to the average person that wants to ask a question and say, like, how's this going to impact me? Simulation runs, okay, here's what might happen. How might I change my behavior with that information?

Bryan Schielke 10:09

Very cool. That's awesome. Appreciate that. All right. Last question here. You operate with values like clarity, willingness, grit, inspiration, and empathy. Which of those has been the most tested as a CEO scaling a deep tech company? And what did you learn from it?

John Cordier 10:24

Clarity number one. So I mentioned that like even part of our mission statement now includes the word clarity, and we take that same definition, having the information to execute with conviction, information, the three parts, intention, context, and impact. So this was a very challenging lesson to learn. There's both like internal clarity as a founder and CEO, internal clarity across an organization, and how do those two things line up? But then that also extends to who we work with because through our software, we're really offering clarity as a solution by running simulations, understanding the impacts that your decisions can make on the populations that you care about. You want to have clarity going into those decisions. And so I try to have the everyone in the company look at our customers and the people we work with as an extension of the impact that we get to have. And that clarity and like how that all of that's connected is really important. And I also think this is an important thing for managers, both in Epistemics and anywhere else. It's like, does your team have clarity in how their day-to-day activities impact key things for the business and therefore impact key things for your customers? Because if you have absolute mission alignment between those things and are clear about it, like that's an organization that people feel good working in. So the clarity piece, definitely the most important value. It's our foundational value, the story of like why we focus on that. And we're trying to build a technology that hasn't existed before, you know, and we're recruiting people from epidemiology, physics, astrophysics, microbiology, economics, like all these different fields. Like everyone starts taking a view of like, oh, well, like if we have this synthetic population of every person in the United States, like here's another angle that we need to consider. And it might be a sociological angle, it might be a psychological angle. It could be something about resource or energy usage and allocation. Well, when everyone starts like taking their own little view at it, it's like, okay, we need to get clear about like what are we really pointing this type of technology to first? And the initial use case was to improve the way that public health could be practiced. So when we understand a public health system, if you start with every individual person, we need to get clear about like, all right, what does every single challenge start with? It starts with the health conditions that we have. So what are the incidents and prevalence of any health condition? Let's let's solve that first. Once you solve that, then the next set of questions is well, what are the behaviors that people are able to take on to adopt different health products and services to improve their health? Then the next layer to that is like, well, how is that paid for? And when you start like layering in these different pieces, you can get more clear about what decisions you can make that actually lead to ultimate goal, advancing health and economic outcomes for a population. So that level of clarity in like the product roadmap was always like a tension between data scientists, engineers, and myself and my two co-founders. Um, and we were like trying to get those groups of people all seeing things eye to eye. And so yeah, I think that the pressing on clarity is the the number one value that's been challenging. But the cool thing about it, it's like I look at it where these values stack. So you once you have clarity, information execute with conviction, like through that conviction, it's like, yeah, I have the willingness to go and try this thing. And once you have the willingness and like you have clarity on it, you do that over and over again. That's basically what grid is. It's like you're gonna keep going, you're not gonna give up, and figure things out, get scrappy as needed. And those are more like intrinsic or internal type values that we're always looking for. The next two are a little bit more externally facing. So we look at the value of inspiration, which we define as show up at your best. It's like if you're able to show up in a really positive way, it enables you to have empathy, which is our fifth value. Because if you're not showing up at your best, like you know, you're not totally in tune with your own self. Are you going to show up and be in tune for the needs of the customers you're gonna work with, whether that's an external client or somebody else on your team? So it kind of goes from clarity to willingness to grit to showing up at your best and inspiration to having empathy. And then once you have empathy with a client, then you're enabling them to show up at their best because they are getting clarity along the way. So it's all connected. And so, yeah, this value of clarity is really important.

Bryan Schielke 14:48

That's fantastic. Uh great example. I appreciate that. That's that's great.