UiPath Fusion 2025 | Shameiz Hemani, Greenlight Consulting LMCU & Chris Ortega, Lake Michigan Credit Union

Clips
More from UiPath Fusion 2025

Shameiz Hemani

CEO

Greenlight Consulting LMCU

Chris Ortega

CIO

Lake Michigan Credit Union

From RPA to reasoning: How LMCU is redefining digital efficiency

Agentic transformation revolutionizes automation by moving beyond rigid, rule-based workflows to intelligence-driven systems. Powered by artificial intelligence agents that can reason, adapt and make dynamic decisions, it enables businesses to automate processes with greater efficiency, flexibility and resilience.Lake Michigan Credit Union has partnered with Greenlight Consulting Solutions Inc. to drive its agentic transformation journey. It’s a strategic move that ensures the institution stays competitive, efficient and member-focused in today’s rapidly evolving digital financial landscape, according to Chris Ortega (pictured, right), chief information officer of LMCU

play_circle_outline Discussion of agentic transformation in automating business processes and improving efficiency.

Info
Transcript

Shameiz Hemani, Greenlight Consulting LMCU & Chris Ortega, Lake Michigan Credit Union

Shameiz Hemani

CEO Greenlight Consulting LMCU

Chris Ortega

CIO Lake Michigan Credit Union

In this UiPath Fusion segment, theCUBE’s Dave Vellante sits down with Shameiz Hemani, CEO of Greenlight Consulting, and Chris Ortega, CIO of Lake Michigan Credit Union, to unpack how agentic AI moves beyond pilots to enterprise-scale outcomes. The discussion centers on “agentifying” end-to-end lending with UiPath Maestro – designing the full journey first, then automating within it – so humans and agents can hand off work seamlessly. Hemani explains why ambitious, outcome-back planning matters (“validate, then fly”) and how focused, Lego-like agents become ea... Read more

explore Keep Exploring

What was the intention behind the modernization efforts at the credit union? add

bolt Powered by CUBE AI

Shameiz Hemani, Greenlight Consulting LMCU & Chris Ortega, Lake Michigan Credit Union

search

>> Hello everyone. Welcome back to theCUBE's live coverage of UiPath Fusion. I'm your host, Rebecca Knight. Alongside my co-host and analyst, Dave Vellante, I would like to welcome two guests to this segment. First we have Chris Ortega, CIO of Lake Michigan Credit Union. Welcome, Chris.

Chris Ortega

>> Hi. Thanks for having me.

>> You're the fun one I think.

Chris Ortega

>> I am the fun one. I'm a hoot. I am a hoot.

>> I'm just kidding. And then we have Shameiz Hemani, CEO, Greenlight Consulting.

Shameiz Hemani

>> How y'all doing?

>> Thank you so much for coming on the show. I'm going to have you both start for the unannointed of telling our viewers a little bit about your organization. Chris, let's start with you.

Chris Ortega

>> Okay. So Lake Michigan Credit Union is a credit union. We're the largest in Michigan, Lake Michigan. And just like a lot of credit unions, started out with teachers, had a good team of people that really cared about the member experience and just the whole credit union just really blew up and we became the largest in Michigan. So we're there for our members. We try to take care of them and they appreciate it. So my whole family's credit union members.

>> Excellent. Now Shameiz, tell us a little bit about Greenlight.

Shameiz Hemani

>> Yeah, Greenlight Consulting is a boutique consulting company. We do a lot of automation and AI work. We spend a ton of time in financial services and credit unions and the team's based in Canada, the US, and Costa Rica.

>> All right.

Dave Vellante

>> So Daniel Dines was just on, he said, "I've pivoted the whole company to agentic." Of course he's been around in automation too.

Shameiz Hemani

>> He has.

Dave Vellante

>> Have you guys done the same?

Shameiz Hemani

>> Yeah, so our view is-

Dave Vellante

>> Describe that.

Shameiz Hemani

>> What's great is Chris has been an ambitious partner in sort of driving change at his organization. And the world of agentic allows us to do things faster, cooler, and way more complicated than we've been able to do before. And Chris's teams have been looking at the world of lending and how to make sure we turned that into an agentified process. Chris, I don't know if you want to talk about that a little bit?

Chris Ortega

>> Yeah, so actually when I started, it was a credit union that was looking to modernize, to really amp up and become leaders in technology in the credit union space. And so it was very lofty goal. And so when I started, there wasn't really much there in terms of automation. And so I met up with Shameiz and we hit it off right off the bat and said, "How do we catapult us into the future? How do we make it so that we are the leaders in technology, that we have credit unions coming to us and saying, 'Hey, how did you guys do this?'" And since it was really kind of open ground for us, it was just made it for a perfect partnership. And Shameiz and I get along really well. We banter off of each other all the time, but we are serious about our work and we get things done. I think we wrap things up in five minutes when we talk. It's like bum-bum-bum. That's it.

Shameiz Hemani

>> Well, one of the ways you described the vision at Lake Michigan Credit Union was ambitious and that is exactly what we heard from Daniel Dines and other leaders at UiPath is that too many companies start too small. They're not ambitious enough. So what is it that really separates the companies that say, "I have this lofty goal, I want to transform how we work," from the companies that say, "That sounds a little risky. I think we're going to go over here and do this little pilot"?

Shameiz Hemani

>> Yeah, I have a version of this I think that's important. One of the ways we do this is, look, you really do want to validate that this stuff works. So we're not asking you to jump in with all three feet, but it is important to think long-term and say, "Hey, we want this process fully automated and agentified by the end of the year. Where do we start? How do we make sure we have that end goal in mind? Let's do a little test here, but once it succeeds, let's fly." And I think that's been the way we go from naked ambition to tempered ambition to reality. And I think Chris's whole company sort of has that same motto.

Chris Ortega

>> Well, for us it's about mindset. There are so many people that are afraid of what AI is going to do, what even modern and emerging technology is going to do. And so for us it's really teaching people what the possibilities are and not exposing them too hard on it right away. Even though that's, like I said, we had an open field to be able to work with Shameiz because there wasn't really anything there. So the people we're getting involved were people who were forward-thinking, wanting to get involved. But the rest of the organizations, they're going, "Oh my god, is this going to take my job? What does this mean for me?" And so we're very careful to make sure that we're going through this organically, that we're teaching, people that we're going slow enough on that end of it, but fast enough with these guys that gets us through the technology. So it's kind of schizophrenic. But as these people are seeing what's being accomplished, they're starting to go, "Oh, okay, so this really works and let's see what we can do. Can we do this? Can we work two business units together from a process that they've been doing manually for decades?"

Dave Vellante

>> So my CIO friends always tell me, "Dave, it's all about change management. Change management. Change management, you really got to really focus on that." My question, Chris, is the way in which you apply change management, is it different with agentic? Is it the same? What are the dimensions that are new that you have to deal with?

Chris Ortega

>> I don't call it change management, I call it transformation. We go in from a transformation viewpoint by saying, "Hey, we can give you a microwave oven instead of a stove." I mean, people used to cook in stoves, but now you can have a microwave oven that makes your job easier. So if I say to you, "Hey Dave, if I could make your job easier and take 25% of your load off of you so you could do other things," wouldn't that be worth talking about, looking at? And when you approach it like that, then they're like, "Yeah, I guess that makes sense," and you show them just something quick and easy for them that takes that process or that task that they've had to repeatedly do. It's a transformation of mindset and it's organic. With change management, it's almost like you're getting them prepared, you're talking to them, but it's almost a push to get them to what you want them to do. This is about winning hearts and minds and that's what we're all about.

Dave Vellante

>> Okay, so that's not different, it sounds like you approach that throughout for any kind of new initiative.

Chris Ortega

>> Absolutely.

Dave Vellante

>> All right. Well maybe that's how you're moving the needle, getting bigger wins.

Chris Ortega

>> And we're doing an agile transformation. We're moving everything to more agile methodologies. And in my mind, you can't do what we're doing without doing agile. It's just impossible. The old methods of old processes from the past, they just don't work. So we are having the whole organization convert to more agile methods.

>> Shameiz, where do you see this in terms of the winning the hearts and minds? So I want to ask you about the messy middle, this implementation state, the implementation phase. When a company has executives on board, they pick their platform, they're ready to go. What is it that derails projects?

Shameiz Hemani

>> I mean, people always derail it. I think we can always get through technology, that's usually not the issue. And so when we talk a little bit about change management, I think the methodology is look, if we're going to make the big changes, let's make them first instead of trickle them down. I think trickling down allows people to get really scared over time. But part of what we did with the home equity line of credit piece is we actually built the entire journey in Maestro. So we did the workflow end-to-end. And at the beginning that's just humans transitioning humans like they used to, but facilitated by an engine, and then we're slowly automating. And how that helps change management is in the end those people don't see anything different. They're just getting documents from an agent now and they're handing it to an agent where before they were just getting it from a person and they're handing it to a person. And so if you can really get that process solidified, understand that end state journey, and then automate within, I think your change management pressure goes away. It's never gone. This stuff's scary, it's big, but I think if you can get to it ahead of time and show them the value of it really is about doing three times more work in a month with less effort.

Chris Ortega

>> So how does that change their jobs though? So they're getting documents from an agent and handing them to an agent, but what about that discretionary 25% of time that they're not doing that manual stuff that Chris talked about?

Shameiz Hemani

>> Yeah, usually they're doing more, but I always look at it like this. We're going to move to a place where the easy yeses and the easy nos are managed by agents and the hard maybes are where we should spend our brains and time. And so if I can sit there and make your job kind of exciting because this is what you do, you try to figure out if this person deserves this loan, let's spend the time there. And you can do that on the 30/40. I think that's great for members. I think that's great for the job you do, and it allows us to be more efficient and scale.

Chris Ortega

>> But that 25%, that's what we find that people will say, "Man, if I had more time, I would do this." But they don't because they're doing this repetitious stuff. And so even with looking through a lot of documentation and applications during the loan process, well now if we can automate that or use AI to be able to make some decision making there, now they got that 25% to do something they've always wanted to do.

Dave Vellante

>> So let's say you take workflow that Shameiz was talking about, you put into Maestro. Have you uncovered situations where, okay, that's the way we have always done it, but there's a better way and agentic is going to allow us to, I imagine Lego blocks, reform the workflow and not only automate it and get 25%, but completely reimagine the workflow and get 10X? I mean I'm making stuff up, but-

Chris Ortega

>> Sometimes. Sometimes workflow is fine, it's just very manual and it has pain points in it. So we identify what those pain points are, where the manual process is, and all we do is just flip it for something that's more agentic base. And then again, that frees up that time for that person to do something.

Dave Vellante

>> What's the practical way in which you go about this? How do you visualize the workflows? What kind of tooling do you use to do that? What's the best practice there?

Shameiz Hemani

>> I'll take it. I'll take it.

Chris Ortega

>> I can take it too.

Shameiz Hemani

>> Go. You start.

Chris Ortega

>> Okay. All right. Because I know that he's going to say something different. So the way that we do is we have our analysts and those are familiar with RPA and stuff, that will go in and we map these processes out with the business unit, say, "Hey, how do you guys do things today?" And we either use UiPath's way of defining that workflow or we write it out, we go on a journey with them and we want to learn what it is that you do and where the pain points are. So then we can then introduce new technology.

Dave Vellante

>> Pen and paper.

Chris Ortega

>> Sometimes it's pen and paper. Yeah, sometimes,

Dave Vellante

>> Yeah, when it's not already in the system. Okay.

Shameiz Hemani

>> I'm not too far off. I think that is the point. The point is to find the journey, the end-to-end journey, look at what it could look like and then put it into MyStore or whatever tools that you guys have. The driver is, in the end, we need to facilitate this process and everyone has to know what this looks like in the end. And where we automate, where we throw agents, where we do it the old-fashioned way, those are like we'll fit into those puzzle blocks. But I think one of the things we all think about is what is my development? How do we deliver this differently than I did when we were doing RPA? And what it's looking like now is, and I think Graham said this on stage, look, building agents as much as I want do more, not that hard. Agents are not hard to build, but what agents need is a lot of context and a lot of data to ground them in the work they're doing. And who knows that best, but the subject matter experts and the business analysts? And so they, I think, become much more prominent in the work we do to deliver these agents. And then we iterate on that prompt and how to make the agent do the work. But in the end, we're grounded in process and context. And that drives the-

Dave Vellante

>> So you can build an agent, I get that, but many of these systems are like single agents or single co-pilots. But as you start to build dozens or hundreds or thousands or hundreds of thousands of agents, is that simple too?

Shameiz Hemani

>> I think Lego's pretty simple, right? You look at the end product and it looks really complex, but the basic building blocks are the same. And I think the idea of we build agents really myopically, they're really focused. You're not building an entire underwriter. You're like, "Hey, there's this one piece of the pie." We build the agent for it, we ground that in context. It's easy to test, it's easy to validate. You stack those, that gets complex and you zoom in, it looks really simple.

Chris Ortega

>> And I love the reaction people have when you define their process or you show the whole process from end-to-end when it crosses businesses, they're like, "Oh my God, we did it this way?" I mean from doing it for so long that they just know that they're that cog in the wheel and they know the whole process, but they know, "Well, my part is this part and so-and-so's part is this part," but then when they see it spread out, they're like, "Oh, my god."

Shameiz Hemani

>> Well, right, because you as in your human doing your job, the workarounds and the tweaks that you've made, adjustments that make the process smooth out, but you maybe overlook the fact that every other human is also doing those similar things that the agent now does.

Chris Ortega

>> Yeah. So that's a way for them to get excited too, because it's like, "Okay, can we make this easier? This looks like a big behemoth mess."

Dave Vellante

>> So we're here at UiPath Fusion where the tagline is agentic AI meets ROI. Talk a little bit about this because a lot of executives remain skeptical that there is going to be some real ROI on AI anytime soon. So what have you seen, Chris?

Chris Ortega

>> So fortunately, our CEO is very forward-thinking and very fast. And she wants us to become a leader in technology. And she is all about bringing as much technology to the organization as possible. So she's actually faster than probably everybody, sometimes even me. So that's been a blessing because she's able to help with the buy-in with the other senior leaders who are not used to that, who do see this sort of a threat even to themselves. And so we try to make it so it's easier for them.

Dave Vellante

>> Let's talk outcomes. What has been the member experience? What's the North Star you're shooting at? Is it more personalization, is it just in and out faster? Describe it.

Chris Ortega

>> We have strategic priorities that we're building for the next year centered around more of the member experience about having loan to approvals process faster. This I talked about earlier, we are trying to make the member experience easier and faster for them, more personalization that we're integrating into our mobile and web experiences. And really UiPath is underlying and functioning at that level that we need them to move things, help things, automate things that brings that data to these UI or user interfaces for them faster so that they can operate faster, they could get their things done.

Dave Vellante

>> Is the data challenge that you're going through now with agentic, was there sort of a JV version with RPA? Is it completely different?

Shameiz Hemani

>> It's a bit different. I do think when we think of data with agents, we're really thinking about context grounding. How do I ensure this agent has the standard operating procedures, the regulatory requirements, my historical data, and any sort of other modes that they need to work in? And historically, we have to go through systems and find this stuff, but today I could take the standard operating procedure from the lending team, make sure it's discrete, maybe updated, and put it into the agent. And I could take the regulatory requirements from the banking acts and the credit union acts and give it to the agent and say, "In the context of this world, make this decision." It's almost allowed us to move away from atomic data to the data we're more used to as humans dealing with.

Dave Vellante

>> So you don't have to go into the bowels of a SAS and rip out the process logic and the process metadata and connect to that?

Shameiz Hemani

>> Sometimes. If we want historical data to be like, "Hey, this is how we've given loans over the last decade, let's just be true to some of this." But more often than not, it's really just about the context of the situation you're making a decision in and you start making better decisions.

Chris Ortega

>> But that context is super important because if the data isn't synced up with the way that context, define that data. So data governance, having that data analytic framework-

Shameiz Hemani

>> 100%....

Chris Ortega

>> is super important. And we have a huge initiative right now because I don't know, credit unions weren't very good at managing their data, neither are banks. So now we're bringing that data together, so we all have an agreement what that data means. It has to mean the same thing for all of us. A member is a member in the same way that you think of it. I think of it and you think of it through a defined dictionary that we put together in our data governance helps to put that together. And that just makes the context of whatever we do with our agentic AI platform that much more robust and more competent.

Dave Vellante

>> And the business context then comes from the domain expert, not the software systems?

Shameiz Hemani

>> Correct. Can we use paradigm shift? I know it's overused, but that is the shift in how we deliver today. All of a sudden that subject matter expertise, the context expertise is kind of the magic. Everything else we can figure out pretty easily.

Dave Vellante

>> And then they'll be tasked with the hard maybes.

Shameiz Hemani

>> Hard maybes. They're the ones that already know.

>> Right?

Shameiz Hemani

>> You're right.

>> Exactly. Exactly.

Dave Vellante

>> Okay. So when there's an exception, humans in the loop, when there's an exception, are we at the point where the agents can learn from the reasoning traces of the humans?

Shameiz Hemani

>> Yeah.

Chris Ortega

>> I would say the opposite.

Shameiz Hemani

>> Yeah, but...

Chris Ortega

>> Yeah, I would say humans are looking at what the AI is coming up with and looking at the reasons why they came up with and then say, "Okay, is that objectively what we were looking for?" And then of course, there's got to be some bias training with humans too to make sure that when they look at that data that the AI is coming out with specifically what they're looking for.

Dave Vellante

>> So it's almost like you're doing reinforcement learning within the credit union.

Chris Ortega

>> Yes, and you may have to re-teach that LLM.

Shameiz Hemani

>> Yeah, jury's out. I think we're going to get there. I think today we're probably at a point where we want to see this work a little bit longer before we sort of bring that back in. But yeah, you're absolutely right. I think we're actually making our processes better, our journeys better, and our people are like, "Why did we do that and do we want to recreate that or do we want to see what happens?"

Chris Ortega

>> And AI is a partnership. It's not having AI take over. I'm a Star Trek AI kind of guy.

Shameiz Hemani

>> Star Trek version of AI?

Chris Ortega

>> Not a Terminator AI kind of guy. So it's all about making sure that the recommendations come from the AI are reliable and that they're learning and using that AI as a partner. And they may override that, and they may have to explain back to the LLM, "Well, this is why I did it." And so that there's context in that context is reinforced. But I see that as always a partnership.

Dave Vellante

>> Excellent. Well, Chris and Shameiz, thank you both so much for coming on theCUBE.

Dave Vellante

>> Thanks, guys.

>> Yeah, this was a fun conversation.

Chris Ortega

>> Great conversation.

Shameiz Hemani

>> Good time.

Dave Vellante

>> Likewise. All right.

>> I'm Rebecca Knight for Dave Vellante. Stay tuned for more of theCUBE's live coverage of UiPath Fusion. You're watching theCUBE, the leader in enterprise tech news and analysis.