Celonis Celosphere 2025 | Monica Chiarini Tremblay PhD, William & Mary & Erin M. Espinosa, Phd, Evident Change

Clips
More from Celonis Celosphere 2025

Erin Espinosa, PhD

Director of Research

Evident Change

Monica Chiarini Tremblay, PhD

Hays T. Watkins Professor of Business

Williams & Mary

play_circle_outline Leveraging Process Intelligence for Enhanced Youth Mental Health: A Year-Long Journey of Data Insights and Pattern Discovery

play_circle_outline Influence on policy through presenting data patterns and outcomes to decision-makers.

play_circle_outline Visualizing System Inefficiencies: Enhancing Stakeholder Understanding in Juvenile Justice and Mental Health Through Process Mining Insights

play_circle_outline Identified issues with repetitive treatment approaches leading to negative outcomes for children.

Info
Transcript

Monica Chiarini Tremblay PhD, William & Mary & Erin M. Espinosa, Phd, Evident Change

Erin Espinosa, PhD

Director of Research Evident Change

Monica Chiarini Tremblay, PhD

Hays T. Watkins Professor of Business Williams & Mary

In this Celosphere 2025 interview from Munich, Erin M. Espinosa, PhD, director of research at Evident Change, and Monica Chiarini Tremblay, PhD, Hays T. Watkins distinguished professor of business at The College of William & Mary, join theCUBE’s Rob Strechay and Savannah Peterson to explore how process mining is being applied to juvenile justice and public mental health systems. Aligned with the event’s “no AI without PI” theme, the guests explain how visualizing end-to-end flows helps practitioners see where system inefficiencies drive negative outcomes, and... Read more

explore Keep Exploring

What progress has been made in analyzing the data since last year? add

What progress has been made in influencing policy regarding the intersection of juvenile justice and mental health data over the past year? add

What factors impact the willingness of stakeholders to accept changes in the juvenile justice system? add

What patterns and surprising findings have emerged from the data regarding the impact of repeated opportunities without consequences on children's behavior? add

bolt Powered by CUBE AI

Monica Chiarini Tremblay PhD, William & Mary & Erin M. Espinosa, Phd, Evident Change

search

Rob Strechay

>> Hello and welcome back to beautiful Celosphere. We're here in Munich, beautiful Bavaria. I've tried to get that right today. I'm excited because we get to follow up on an interview we did last year to really into some stuff that's, I think, really important work that's being done, and that really always talks to what Celonis is about from a cultural perspective as well. Right now, I'm excited to welcome back Dr. Monica Chiarini, almost, Tremblay, who is the Hays T. Watkins professor of business at William & Mary. Thank you for joining us.

Monica Chiarini Tremblay, PhD

>> Thank you for having me.

Rob Strechay

>> And Dr. Erin M. Espinosa, who's the director of research at Evident Change. Thank you for your first time onboard.

Erin Espinosa, PhD

>> Hey, thank you. Thank you for getting the M in there.

Rob Strechay

>> I did. I saw it there, so I thought it was important to put the M in there. We like to be very, very formal here. Now, for the informal part of it. I think, again, we're looking back to last year and everything that's been going on. You two have really joined forces to bring a use case for applying process intelligence to really improve outcomes for youth and from a mental health perspective in particular. Why don't you help us understand where we are on the project status and how things have been going?

Monica Chiarini Tremblay, PhD

>> So, it's interesting, at the last year when we last spoke, we were just getting started. We had ingested the data, we were starting to look for patterns, but we didn't quite know what to do with it. And so, what we've done a lot of this year is to just get in front of people, present it in front of academic audiences, present it in front of judges, present it in front of stakeholders to see what the possibilities are with this type of approach and what other context might even work with this approach, let's say from a healthcare perspective. But I'd like Erin to add, because she's actually on the ground.

Erin Espinosa, PhD

>> Yeah.

Rob Strechay

>> You're doing the real work.

Erin Espinosa, PhD

>> The real work. I have a lot of fun. I will say this, I think our partnership has been made in the purpose of making things better for children and families across the world. And what process mining has brought to the game is the ability to tell systems in a different way how their inefficiencies are actually having negative outcomes. And if we hadn't had this opportunity, I never even would've thought to use process mining. I would've been stuck with the traditional binary logistic regression and standard research procedures that practitioners don't understand. And I think one of the things that's been real powerful the last year is being able to take an actual visual of how kids flow through these systems and show judges and probation officers and clinicians that there is an impact when you make a decision that goes in the wrong direction.

Rob Strechay

>> And I mean, I think, to me, I love this when we were talking about it last year because it's so important, because I think, again, you want people to get better. You want the system to work for them, not against them when they go into it and have positive outcomes. Monica, you were talking last year about influencing, really, policy. And so, how has this really come about in the past year? Has there been any progress on the policy side of things or where do things stand?

Monica Chiarini Tremblay, PhD

>> So, we are still learning how to present this information. So, you can imagine a data set has what, 700,000 kids, 450,000 of them that intersect with mental health. So, juvenile justice data, mental health data, 458,000 of them have data in both data sets. And we're trying to figure out how to interpret these patterns because you can imagine not one kid follows the same process. So, what are the key performance indicators in this case? What do we measure? How do we communicate that? And the only way for us to gather that information is to get in front of people and say, "This is what we're finding."

Rob Strechay

>> Yeah. To me, that ties back to what you were saying about some of the analytics and what we would call traditional ai, the regression analysis and other things and using that. But when you're looking at the process intelligence. How does it really, I guess you could say in your experience, Erin, bridge the gap between the process intelligence and real life and what's really happening?

Erin Espinosa, PhD

>> Well, what I can tell you is in my experience being a former probation officer, working with these children, running a state juvenile justice system, now being responsible for a large research firm, what I can tell you is that until you can actually sit down with folks and help them know that you truly understand how their system operates and how they operate in it, and then how the children, such as Sue and Sam and others experience that system, they're not ready to accept that there's anything to change. And what the power of process mining has allowed us to do is almost in a real-time way, show them that in a much better fashion than the big marker and crayon, which is the way I used to do it, is just drawing it out and hoping people like my doodles. But it's pretty powerful stuff when you're actually able to show in an aggregate, literally kids spiraling in and out of detention and trying to get treatment and not getting the appropriate treatment and going off into prison-

Monica Chiarini Tremblay, PhD

>> Waiting forever to get treatment.

Erin Espinosa, PhD

>> Hundreds of days, even. And what's even more fascinating is it started conversations in places like district attorney's offices and judges and courts and probation systems around inefficiencies and how inefficiencies can lead to bad outcomes, especially when we allow those inefficiencies to continue to match to inappropriate care for a very long time and don't have the ability to pivot really quickly.

Monica Chiarini Tremblay, PhD

>> I would say probably one of the key things is sometimes some of the people that are real experienced in the system... Like we presented this to a judge and she said, "We know this. We've known this." I'm like, "Yeah, but did you have evidence? And what have you done about it?" And this provides us the ability to give them evidence. This is real kids that were showing you and this is their data.

Erin Espinosa, PhD

>> Yeah, last year we started our talk out talking about Chris and I have a name for almost every scenario I can think of of the families I've partnered with. This takes Chris Times 400,000 and helps to show the whole community, whole organization, states, counties, territories, even, how that compiles into negative outcomes. And so, it takes Chris, not out of the equation, but gives his voice a much larger one even.

Rob Strechay

>> Right. And you can start to see the patterns and things like that. What's been surprising that's popped out of the data that, again... Because I look at it and we've talked to a lot of people. And have you been able to influence changing the process? Hey, we know if it goes down this path, it's going to be one of those do loops where you're coming back over and over again?

Monica Chiarini Tremblay, PhD

>> One day I was getting ready to give a presentation at IE at University of Madrid. So, I was playing with a dashboard because I knew that there was going to be folks from public health and from the juvenile justice system in Spain there. And one of the things I found is that kids that are given lots of chances but with no repercussions, actually just get stuck if they're just constantly getting put on probation, but there's no intervention. And when I presented this, they said, "We see this in Spain as well." And for you maybe it wasn't like a wow, because I always think of no, we should not be punitive.

Rob Strechay

>> Right.

Erin Espinosa, PhD

>> Actually, I saw it from a different perspective. I see it as a true test of what I call, I cannot deny the null hypothesis. The fact is that in this particular data set, the kids were limited to the type of treatment that was approved in the Medicaid state plan. And so, you can only do so much of the same thing over and over again and it doesn't work anymore. And so, those children that Monica is describing that we called a detention spiral, what you could see is that they were coming in for low-level offenses, but going right back to public mental health and getting the exact same treatment over and over and over again, we were modifying the approach and then you would see them skyrocket straight into the prison system.

Monica Chiarini Tremblay, PhD

>> I love the example that you give, Erin. Sometimes it's like you show up at the emergency room and you're like, "I'm having a heart attack," and they're like, "Here's an aspirin."

Erin Espinosa, PhD

>> Yeah, that's literally it. And to be able to see that with hundreds of thousands of children has a much bigger impact than me just saying, "Hey, there was this kid once that you worked with." It carries a lot more weight.

Rob Strechay

>> And I would expect because companies have this problem in spades, but governments in general and just knowing some friends who are in government, the siloed of the data and getting access to the... I know you had this dataset and you were getting it all loaded up last year and all of that, but do you see that these agencies... Because it goes cross-agency. And has that changed or have you been able to change the mind of people that, "Hey, you keep your data as the system of record, but we need to un-silo the outcome of the data"?

Monica Chiarini Tremblay, PhD

>> It's a social technical problem because data sharing is not an issue, it's the people that don't want to share, and sometimes for legitimate reasons. They're concerned about the use of that data. But it can be very powerful, as we have demonstrated just with this small, just two agencies. Imagine if we could get education in there too.

Erin Espinosa, PhD

>> Or child welfare. I think part of the fear though is that when you're dealing with public systems and public systems data becomes public, then it can make workers fearful that that data will then be weaponized against their jobs and not truly coming back to say, "We want to do our jobs better and that we want to keep families and children healthy." And you see that a lot. So, I think part of it is also understanding that data in a way that isn't making people scared to want to talk about it and then being able to have some actual steps to do something with it.

Rob Strechay

>> Yeah, I mean, to me, that's got to be just beyond the regulations and everything else and all the stuff that's coming down the pike from an AI perspective and things of that nature, it would seem that this is a place where there hasn't been a lot of this process analysis gone on. I mean, the process re-engineering's been around for decades everywhere else, but are you seeing that people are becoming more receptive of it? You can show them the actual picture of it. A picture is worth 1,000 words in a lot of places.

Erin Espinosa, PhD

>> That's actually how this started. When I work with large governmental systems, juvenile justice and mental health, primarily, what I learned is if I didn't actually draw out how they experienced the work, literally through a color-coded map, they didn't believe me when I was telling them what their data was saying. So, when Monica brought up the idea of using process mining as an opportunity that with administrative data, it fit the mold perfectly. What did I miss there?

Monica Chiarini Tremblay, PhD

>> No, your diagram was beautiful. Actually, the colors would merge and blue and yellow would become green. I mean it was gorgeous, but I'm like-

Rob Strechay

>> She's an artist, a process artist.

Erin Espinosa, PhD

>> My staff are the artist. I just had the idea.

Monica Chiarini Tremblay, PhD

>> But I was like, "There's literally software that does this, Erin. Let's put it in the software." And we just learned that we published inform journal on computing, but using machine learning techniques. It takes that long to get publications out. But I'm like, "This is so much better than a static machine learning algorithm because it's got timestamps and this is a perfect tool."

Rob Strechay

>> Right. And because it has that time-boundness to it from that perspective.

Erin Espinosa, PhD

>> Well, not only the time, what we call the path analysis and how many kids are going through. And what I want to add to that is that you have more organizations nationally that are pushing for what they call sharing of data sharing of case management software right now. But you don't know what to do with that data. But with what we're working on here, we're actually able to show at what intercept in the system a decision could be made differently or a intervention could be held differently, and then that path could change. That has a lot of power when you get in front of the right people. At least we're hopeful because that's what we're doing this next year.

Monica Chiarini Tremblay, PhD

>> Right. And we hope that this can be a pilot to then go bigger, ask for more datasets, merge more information, like child welfare, because I'm sure that there's a intersection with foster care, for sure.

Erin Espinosa, PhD

>> Oh, yeah.

Rob Strechay

>> Yeah. So, I mean talking about the next generation and where this goes, I mean obviously at William & Mary and you're training the next set of analysts that are looking into this stuff. Is this tying back into that and helping them see the potential?

Monica Chiarini Tremblay, PhD

>> So, I've actually brought quite a few graduate students because this project is so large that you can take many angles. So, we are doing an economic analysis. And in that analysis, we've involved an undergraduate student, we've got several graduate students working on looking at the path analysis and deciding what a key performance indicator is. So, I've got a pretty large group of graduate students now looking at it. And I like the aspect, William & Mary's a liberal arts institution of data science for good.

Erin Espinosa, PhD

>> I love the pathway of students who are going through the business schools actually getting to see how their skills can be applied to change social justice systems. And remember, these are systems that our tax dollars go into anyway. So, I feel like we have to start challenging them like we challenge Amazon and others. Do a better job delivering what you're set out to do, and we can't do that if we don't empower the youth coming up to have that voice. Is that fair to say?

Monica Chiarini Tremblay, PhD

>> I would say, yeah, they're not even aware of these problems, right? They are young. And it's so good to expose them to it and to empower them to try to make a difference.

Rob Strechay

>> Yeah, I know because both my kids are at university now, so I understand how they think, fortunately or unfortunately. And to your point, I always tell them, I'm like, "Go out and see beyond this and try stuff outside your area, so that you become aware of more that's going on because that's so important." So, final question. Hopefully, we'll be here again next year and we can do another update. We'll start with you, Erin. What do you hope to say next year at Celosphere 2026 that you can't say today?

Erin Espinosa, PhD

>> I would love for us to actually have a place where we can demonstrate this actively occurring in a system that's trying to reform itself.

Monica Chiarini Tremblay, PhD

>> I would agree with Erin. I would love to have not a pilot anymore, but an actual project.

Erin Espinosa, PhD

>> Yeah, I don't like to call things pilots because it pilots stay around forever. Demonstration. We're going to demonstrate it can work.

Monica Chiarini Tremblay, PhD

>> No, no. I want it to be want be an actual project-

Erin Espinosa, PhD

>> Oh, done. It's installed....

Monica Chiarini Tremblay, PhD

>> where the municipality-

Rob Strechay

>> To your point, that's why even at Amazon, they used the word preview because it goes from preview into GA.

Erin Espinosa, PhD

>> Right. We want to see it previewed.

Rob Strechay

>> Yeah, you want to go from preview to GA, and that'll be great. So, I look forward to being able to talk about this. Thank you for coming onboard and giving us an update. It's really important stuff and I thank you for doing it.

Monica Chiarini Tremblay, PhD

>> Thank you for having us.

Erin Espinosa, PhD

>> We really appreciate it.

Rob Strechay

>> And thank you for watching this episode from Celosphere, live here in Munich 2025. We'll be back with more on theCUBE, the leader in analysis and news. Stay tuned.