PegaWorld 2026: Supercharging Sales: InterSystems Journey with Pega AI

Thanks everyone for joining. We're here with the InterSystems team, joined by Kanishk and Karri, who are some great colleagues that I started working with over the last several months. And they're going to tell a really interesting story about how they have been using Pega and Pega Sales Automation and really modernizing what they're doing. And I think there's a lot of great insights about how they've been going along that journey and how they're really leveraging AI and AI agents to make their CRM experience really, really powerful for their end users. And I think that's something that you guys do really well, is put a focus on the end user experience over almost anything else. I think that stood out to me. And you guys will do a little intro in a moment, but I did just want to tell a little story about my first couple interactions with, with you guys.

And, um, It started, I think it was last fall, and I got an email saying that Corey and Nish really wanted to dive into agents and coaches more. Um, I said okay. So I had been working kind of the origin story of Pega agents came from Sales Automation, which I don't know if you guys know, but I was, I meet with Alan. He's probably my most important user. Um, and I meet with him often and he was explaining a use case and he said, whenever I look at an opportunity in Sales Automation, I always ask the same set of questions and I run it through the same sort of, um, things in my head. And he's like, well, what if we could combine that or harvest that, put that into sales automation, and then everyone could take advantage of it, right? And that is how Pega agents were born.

And so I got a, we basically built that custom and started adopting it. So I kind of had this head start of adopting Pega Agents and coaches, and it was kind of like this little superpower that I had a secret. And then I got a phone with Corey and Nick, and Corey immediately started peppering me with all these questions about how to use agents and coaches, and it was really powerful, and I felt like you immediately saw the value and it just hit me that you, you guys both were just these great technologists really outside of anything else. And I think you guys are going to tell that story of how you got to where you are today. So that was just something that stuck out to me when I was thinking about this session. Yeah. Um, and so if you guys want to do a little intro of who you are and what you guys do, that'd be great.

It was a very kind words, very generous. Thank you. Um, I'm, I'm a principal architect at InterSystems. I'm on Corey's team. We own the sales tech and the BI technology stack. Yeah, I mean, that's pretty straightforward. But to Andrew's point, the minute we got Ahold of Pega AI and from a sales automation point of view, it was this mindset shift that kind of happened. It hit me all at once in this aha moment, kept on happening over and over again. And once I'm like, please talk to our account executive, like will get me the product guy. Who can I talk to? Jump on the call and had like a thousand questions and technologist is the point there. I mean, it just opened up the door for so many possibilities. We'll cover it today and I'll do some demoing with you. Yeah. Um, but overall, I think it's just, it's really just this mindset shift and I'm trying to hopefully for you guys in the crowd, like understand and try to relay that to you so you can kind of use it to your advantage and, and how you use CRM and how you use Pega.

So, and also talk probably about some lessons learned. If you guys were to do anything differently, what would that be? And your thought process for all of these? I think everyone probably has a million AI use cases that they could start with. And I want to learn from you guys. How are you figuring out where to go next and what to prioritize and your thought process there? So I think we'll get into all of that and then see a cool demo, some real stuff of Corey showing what he can build. Yeah. And hopefully it's interactive and I want to get some feedback from everyone in the crowd and the crowd. I say we have a full audience here, by the way. It's just packed. Unbelievable. So thank you all for coming. Let's jump into it. Okay, perfect. So we'll start with the short video. Every data company makes the same claims.

Security. Integration. Scalability. InterSystems isn't just another data company. We're different by design. We've spent nearly 50 years building solutions that meet the unique needs of our customers, delivering reliable, unified data when they need it most. Starting with our unique architecture designed with speed and accuracy in mind. We've brought the processing closer to the data without having to be moved or duplicated. Data can now be processed faster with fewer errors. Our software architecture delivers unparalleled speed to our customers, so they can think fast and make quick decisions with confidence, all based on actionable, error free information. We know our partners need to be able to leverage different applications and data sources in order to work successfully. That's why our software design priorities natively process any kind of data seamlessly, from hospital systems to financial trading to space exploration. We're proud to be the platform of choice for the world's most critical applications.

We've spent half a century working with complex data and delivering reliable solutions for our partners, constantly evolving and innovating to ensure you have the best possible solutions so you can be ready for whatever comes next. And that's not just what sets us apart from other platforms. It's what gets us out of bed in the morning. InterSystems already ready. Excellent. So that's a little bit about us to give you some more context, where a database company big in the healthcare sector globally, about a billion patients is data sits on our database, all of epic, all of the VA a lot of the NHS, Saudi Arabia, we have a global presence and this company was started about 50 years ago by one person at MIT. He was doing research at MGH, and he saw the potential that the technology had, and he turned it into this mammoth.

And he still runs the company. He is our president. And what that means is that it's a very unique company. We've never raised venture capital, we've never gone public. And we try to meet our customers where they are. We meet doctors where they are, we meet clinicians where they are. And that means that we're in the Or trying to understand the problems they have so that we can solve that for them. The implication of that is that a rigid sales process just does not work for us. And that's really why we made the transition to Pega in the first place, because Sales Automation is based on your low code automation system, and that had a lot of value for us. It allowed us to kind of structure the processes, the way it worked for our sales rep, the way it worked for our regions. And more than that, what's really powerful about the low code layer is that you can have non-technical users also contribute, and we have a pretty big team of both technical and non-technical users, and everyone is able to contribute, resolve issues, figure out what's happening.

And that's really where the power of low code lies for us. Hopefully that keeps increasing as yes. Absolutely exposed to Infinity studio Infinity Blueprint. Right. I mean, with with AI, it's become so much more easy to add on new functionalities. Figure out any any issues. So, so talking a little bit more about our journey, right? We started out with, say, we migrated our CRM to Sales Automation and we unified our go to market on a single platform. How many years ago was that? We started about seven years ago, right? Seven years ago. And I won't go too off topic, but it started with just one machine. I actually downloaded and installed Pega on our on prem environment. We needed to be on prem. Pega offered that option for us. We were able to build it in our own system, and it kind of went off from there.

So it was very it's a it's a very humble beginning. Great. Yeah. And then the system obviously grew, right as people saw the potential. As we drove adoption, the system grew and the standardization of Pega kind of created the starting point for our GTM journey. And then post that, we realized that we really needed to be on the cloud. That realization came four years ago when I was at PegaWorld, and they were just starting to demo some of the AI technologies, and I think it was incredibly clear to us that being on the cloud was the way to go. Scaling a GTM platform on prem, especially something as complex as Pega on the back end, was just not feasible. And when did you complete your cloud migration? About a year and a half ago, I would say. Yeah. I think it's two years. Two years ago almost maybe.

Yeah. I mean, as as Cody mentioned, we went with Pega initially because they offered an on prem instance, because we're very, we like, we're secure company. We like having our systems on prem. Healthcare in general is one of those industries where people like having end to end ownership of everything. But we met with the Pega security experts, the CISO, Carlos, and he just blew us away. And we came away assured that our data was secure. Yeah, that was an excellent meeting. It was between Ken Morrison and Carlos and a few of our executives, and meeting with Pega was fantastic. They we actually had a couple workshops after that to try to help each other with security on the cloud. And I know at the time, even like four years ago, people were very scared to go to the cloud. They were worried about their data.

And now it seems to even be even more secure there. So that was fantastic. Yeah. And Pega was the first internal tool we had where we moved it to the cloud. That kind of set the standard for every other migration we've had since, because up until that point, absolutely nothing was on the cloud. We actually had our SharePoint on prem as well. Believe it or not, it wasn't on Microsoft servers until a couple of years ago. Good call. Kanishk Mittal. So the cloud modernize the foundation, right? It kind of gave us it gave us the the infrastructure we need to start really, really scaling this. And then over the past year, we've really dived into Constellation. And this matters because if if you're on a sales org, you know that there's always friction between the CRM team and the salespeople. The salespeople think, hey, this is a waste of my time.

I have to put stuff in just to kind of appease management so they can track what's happening. And they're always hesitant, and they only put in stuff at the last minute when an opportunity either fails or has just gone through. And with Constellation, this is the first time where we've noticed that really changing. You know, people think that the system is intuitive and and this is across different sectors, across different age groups, that we have a lot of users who haven't been, who aren't as computer savvy in the nicest way I can say it, but even they find Constellation to be super intuitive and the amount of inbound requests we have, questions we have in terms of issues while putting data in, resolving data, closing leads, they've just started to go down drastically. And obviously it's going to touch on the next part, which is AI.

It's a combination of a much slicker, easier to navigate UI on top of this foundation, this work we'll talk about with AI. So I think it really started pushing people and getting them almost interested in wanting to be in the CRM and get value out of it very quickly. And I think this, I mean, this sequence is very important, but was this a planned sequence or did this just kind of happen organically? It was fully organic. I think every PegaWorld I come, I see something new and then I go back and I'm like, hey guys, we really need to do this. He comes over, he pushes me. He's like, let's do it. We need to, we need to start doing this. I'm like, well, let me let's hear it out. We work on change management. And of course, if anybody has questions on change management at the end, please let us know because we've had to deal with maybe over 500 sales users across 25 different countries.

And the transition has been. Good. Context. And, you know, I'll talk about it. It is funny because internally we obviously use Pega Sales Automation, and we went through literally the exact same steps, right? We built sales automation and then we were on prem for a long time, went to the cloud, adopted Constellation, and now it's using the agents. And it's been, I think the same thing you guys just said. We're finally seeing people enjoy using a CRM app, which for anyone who uses CRM, knows that's literally impossible. Um, and it's real. It's a real shift. And people, I ran into a, one of our sellers in the Innovation Hub and he came up to me and said, yeah, I mean, there's nothing I can complain about anymore. So yeah. That's good to hear. And it's cool that you guys are validating that as well. I think you touched.

You know, not to go back, but like one of the reasons we went with you guys as well, you're very similar culture with your sales organization, the way you're set up. You were across the street from us for a long time in Cambridge, mass. Right. And, um, the culture just made sense. So your transition and, you know, we're in the same offices around the world, like offices, same countries around the world, and kind of same set up. And, you know, from MIT and chess playing, you know, genius guys. Yeah. Dartmouth guys that are, um, you know, it just made sense. So I always like to put that part in because even though the technology is great, the partnership is fantastic. And I think it shows that I'm able to have a great conversation with you over a teams call. Right. And I think, you know, we obviously are using Sales Automation, but I think you could remove that line.

And any other Pega customer would probably have a very similar journey, to be honest with you, starting with Pega Platform migrating to the cloud, adopting Constellation, and now hopefully taking advantage of the AI tools. So this should be applicable to anyone else who maybe uses Sales Automation or not. Again, a lot of it also just comes from the fact that we can talk to people like you, write with other vendors, you know, you're not able to talk to protected resources, you're not able to talk to PMS or developers. And the fact that we can, we can reach out to our AI. And he just gets you on a call within a day is just so phenomenal, right? Because it shows us that you want to work with us, that you're open to taking our feedback and it helps us resolve any issues we have so fast.

We could get quite annoying sometimes, probably for. The apologize. Folks. They all know who we are at this point, I think. Yeah, no, but it's been it's yeah, it's great because I think and anyone else in the room who have use cases, we should talk about them because what you guys come with is directly going into the platform and making the platform better for any user of Pega. Again, you know, Corey was pushing us really hard on getting MCP servers natively integrated into our agents. And he's like, I need my MCP servers. Like, okay, we're working on it. And so we got that built and now I've been using it and now you're going to be able to use it. And everyone else who uses Pega Agents can leverage a really good MCP experience. So I think we need people, customers like you guys who are pushing the boundaries, taking the latest stuff and seeing what it can do, and that'll help everyone else who uses the platform.

Excellent. And talking a little bit more about the value, right. And that's where AI really comes in. Right? I think the chief complaint that anyone has with the CRM is, hey, I put things in, I put time and I put data in, what am I getting out of it? Right. And historically, the answer has been absolutely nothing. Right? But what we tell people now is, hey, if all of you are pulling in your information together, you can collectively leverage that hive mind, right? If you're trying to work on opportunity on a particular product, and someone else has done that and put their data in, or if you did that three years ago and you put your data in, you now don't need to go back to your email, to your Excel files, to your word docs. You can just go to Pega AI, talk to the opportunity and say, hey, you know, what are the next best steps?

What happened last time? And it comes back and gives you suggestion. It gives you all the context you need. And not that I know. We're slowly going the slides and we'll pick it up a little bit. But a use case that came up. We usually have to email and it's probably happens at all your companies. Hey, I'm trying to sell into this market. Has anybody had any luck here. They just email world wide sales. They email everyone. They try to send these emails out. I took the exact email that the one of our sales reps took put into Pega, AI and verbatim put it back into the world wide sales chat. I got about 15 to 20 people reach out to me personally say this blew my mind. And that's just the tip of the iceberg, right? And if you can keep on doing that, imagine how much time people can save.

And people want to put more data into the system. They want to put more data in. And this is, you know, we're diving into change management a little bit, but but the way we've done Pega AI is we have a de facto enabled it for everyone. We have responded to emails with, with, with responses from Pega, AI. And people reach out to us and they're like, hey, we want access to this. So at this point, I would say we have about 30 to 40% of our sales force. About 260 people enabled for the beta. So we're still in beta for Pega AI. That's a very large beta. And they've all requested it themselves and they're using it. That's like half your user base. So that's that's great. And I think, I think one thing maybe we glossed over, but just to make sure it's clear is you guys saw right away the value of it is having the Pega cases as context, right?

So you're passing in all this data, you're having a conversation with Pega AI, but it understands what you're talking about, right? You're not just going to Claude and saying, how can I make this opportunity better? And then you have to re-explain everything is taking advantage of the entire data model. And Pega AI is just built natively right in there. Actually go to the next slide. It will I think. That was our journey, you know, modernization, adoption, and now we're intelligence. So yeah. Um. So let's talk a little bit more about the sales assistants, right? And we talk about what it does and what the impact is. So it can surface relevant data insights in real time, right? Your sales manager, you have ten different AI supporting up to you, and you want to figure out what's happening with each deal, right? Traditionally, they put all of that into an Excel file or they write down their notes.

And now you can just put that in Pega as an activity update and the manager can go in during their one on one and say, okay, show me all the deals that have been updated in the last week that haven't been updated, that have basically died out where there's no activity. And the impact is that you can spend a lot more time selling instead of searching for things, updating things. So all the manual work that they've had to do was just evaporated. Selling and probably coaching to the manager can do what they're best at, which is helping the reps. Yeah, just chasing them for updates. You can ask the agent and someone, a few of our managers have already done this. I'm meeting with this person today. What should we talk about? Just that's the prompt. Yeah. And since we have all the context from our CRM and we'll touch on it a little bit, not only servicing data, but also entering data.

And if both ways go a reduced by 80 to 90%, and that's literally the number, then you could take that 90% time and really start coaching, reviewing the pipeline and more. So I'm not going to skip ahead. But and talking about the MCP, you know, the next step we want to do is we want to integrate all our outlook data into Pega itself. So instead of people having to manually add teams, meeting transcripts or outlook invites or emails. All of that context will be there. So it's going to be this, this supercharged AI system where you just go in and you're like, I'm meeting with this person. What do I do next? What did we last speak about? What are the other deals that we worked on that are similar to this in terms of the product, the customer size, what should I say? Yeah, save all that time.

But even if you think about it that way, it's like copilot what you're saying, but it's copilot with our entire company's system behind it. Yeah, really. So all the opportunities that happen, all of our products, billings licenses, and I'll show it quickly in the number of knowledge sources we have. It's I'm trying to tie. I'm a data monger at the at the moment, I'm going to all the different silos in our company and saying, we need your data because watch what can happen and we trade. Give me the data. I'll give you some. What is that done? Because we've we've seen this does it. Sometimes it shines an uncomfortable flashlight on maybe bad data. Does that happen to you guys? Oh, 100%. Sometimes. But people quickly change their mind on that. What happens is they're hesitant at first to release all their data or, you know, there's a silo for a reason.

Your your data, you have ownership you like, you really trust how you analyze the data. But once we put it all together collectively with everything else and we start asking questions, even people in systems that have never thought would come over to our system, they're like, we need to expand this to more people in our department because we're saving so much time before we'd have to send emails to five salespeople and do a circular thing, have a meeting on something, just go right to the agent. That's the point. Yeah. You eliminate all the searching across systems because you bring it in and this accelerates the cycles, right? So you have your sales enablement system or you have your enrichment system, right? And what's nice about Pega is that you can integrate with these systems in a day, right? You have the service wizard, you can basically paste a sample Json and it'll build out the integration for you.

And in a day, you can have two systems fully integrated without any manual work. And now the agent can consume that. Get that data in as context. We don't have to, you know, we can go into Blueprint later if we even have time. The next piece of the journey. That's the next. Piece. That's next. Year. That's next year. But actually. But we even connect to our on prem iris database that runs on our technology, that runs a team. It's CRM, BI, so CRM BI. So it has all the data. We can actually Jdbc right from Pega query it live with with the with cloud, run it through an LLM without servicing any of our data in the LLM through this masking process and this gateway service that you have. And it's just, it's fantastic. It's pulling up data that's not even inside Pega and surfacing it and analyzing it like it is.

Yeah. And you can use that to figure out what the next best action is, right? If you want to look at billings data, licensing data, and this sort of improves your win rate quite a bit because you're so much better prepared for a conversation, you know, you're going in armored with all the context that you need for that customer, for similar customers, for products that have solved similar problems for customers. And you just don't have to worry about any of that. And it's context aware, right? It knows where the rep is in the cycle, whether this is early or this is late, whether you need to bring in a C-suite person or you need a CHAMPVA, it can figure that out because it runs off the metric framework or any framework of your choice, really. And so there's no separate tool here, right? It's just this one chat bot, which has everything, right?

We've built out everything on the backend and the reps just go in single chatbot interface and they get everything they need. And you have all the reps using the same tool, which I think is that's really important. Besides, the other scenario would be people are logging into all of their favorite AIS and doing whatever it is they want to do. Now you've got a tool that's really powerful, and it probably has all of your best practices baked into it. 100%. Yeah. Cool. Talking about the knowledge, right? So it captures every deal and it accelerates it. Every rep moves faster and the system knows what works. It surfaces the insights, the intelligence and effective messaging. And as you said, it standardizes it, right? Everyone in the age of AI is a builder and we love that. We appreciate that, especially as a technology company. But at the same time, there is value in standardization, especially with sales.

You can scale it, right? It's not trapped in silos anymore. You can share it, you can anonymize it. And that's something we'll talk about a little bit more as Cody's doing the demo right. That Pega AI is super secure because it's baked into the SSO privileges that different users have. So you don't actually have to worry about unauthenticated access in any way. And that's really big, right? If you do this outside of your CRM, you have to manually go and set up user groups and privileges and whatnot. So that's. Big. I'm sure you'll get into it, but the data that you pass is all controlled by our rules and access based. Yeah. The agent will literally say, sorry, you don't have access to that data. We try to be as open as possible. And that's kind of our practice. But sometimes there's sensitive data that not everyone should be able to see.

Yeah. Yeah. And it improves every quarter. It's not static. Right. Because as you can see, the more data you put in, the better the responses will be, the more context you'll have. And historically, we may not have had the best data because as we spoke about, the sales reps kind of have this, this trepidation where they're like, oh, I'm putting in data, but I'm not getting anything out of it. Now they're putting in everything they want to put in everything, even if it's not as relevant to the deal. They just put it in the notes because they're like, oh, I may want to query this later. Right. Yeah. Great. Great. And it learns. It improves the volume compounds over time and scales. So let's move on to the demo a little bit. Oh yeah. Do you want to just go right to the demo. I mean yeah.

We can we can come back to this. But yeah, let's let's do that a little bit tight on time. Yeah. Cody can start the demo. I'll walk. The demo. Yeah. Perfect. As you're getting that set up, I just I was curious for you guys. I know you recently traveled and you went and got a lot of feedback. That's something that we struggle with. Like how are you prioritizing? Well, first of all, how are you capturing feedback? So I know you've got a pilot going, but how are you learning or seeing what people are saying about your. The workshops really, and I call them workshops less than trainings. I mean, it's really paid off because essentially we get the time. People are like, oh, what's the agenda? What's the it all dependent on the people? So the idea is that I showed them the tools. We do a short demo, go through the kind of foundational tools we have and say, what would you think?

What would you like the system to do? What is going to help you do your go through your day faster? What's going to help you cut down that administrative work drastically? And then we take those and we and the idea, I think the biggest part is, is a foundation. What I'm trying to do to prioritize those is anything that's going to contribute to the overall foundation of the future of where we're trying to go. Those get a higher priority. And even though there's more particular use cases that are very niche, or those will all be solved in time by building a very strong foundation. Yeah, maybe a bedrock, a bad a bad joke, but a bedrock, a foundation. And you'll explain when you say foundation, I think you mean your set of tools, right? Yes. So this is a I'm sure some of you have seen Constellation.

This is just our this is our, our pretty much our development environment. So you see some test stuff going on. so that's fine. Um, and really what I meant by an agent. The CRM agent, we literally call it that the CRM, CRM agent. And what sales could do is they could pretty much go here. And if they're new to the organization, obviously it does self documentation. Every time you build a tool, I can just run this prompt and say, hey, tell me what's going on, what's available, what's been added in the last couple of weeks. I don't have to build any documentation or anything like that. So of course it knows who I am. Each person in our system has a profile that's built in the context, and I'll show you that in the back end. But essentially what we could do with this agent is a lot.

So search for orgs, leads, contacts, ops operators, sales activities, tasks, marketing activities. We tied all of our marketing activities into the system. Uh, WRC is actually our, one of our systems that we connect to a support system. Okay. We're not currently using Pega as a support. We're homegrown support system that we use our own technology for. Could change that. But which works very well. But we're able to, like I said, Jdbc and get all that rich. This list is incredible and this is what I'm talking about when you bring me with questions. I mean, this and. And right now we can essentially and I'll show you, but you can dive in to each of the particular cases, contacts, orgs, ops leads, and leverage them all within one space. And to just add on to that. I don't know if you can see, but there's a small attachment option there, right?

And historically, sales apps have come to us and said, hey, I did this bulk update. I've done this stuff in Excel. How do I get this data in automatically? And we're like, well, you know, do you really want to do that? We don't want to build out that functionality because then you have garbage going in. But here they can do that. And the AI flags issues that it says, oh, hey, this whole field isn't valid. You need to update that. Yeah. Um, you know, I don't have a meeting or like I'd say even I use my phone. You could take a picture of a business card. You could take a picture of the whiteboard and say, help me understand this whiteboard has a context of the entire CRM, even with just like chicken scratch and like abbreviations, It completely understood everything they're doing. It solves a real user experience challenge, right?

Like you don't have to know where to go to create a lead or a task. You can just talk to your assistant and it knows that it's. A single pane. Yeah. And it's going to help you do it. Yes. And it'll. Help that. Before it was data entry. You'd have to like, okay, I want to build a lead. Like you can say, um, what are my open tasks? So again at the top, and while that's kind of given a response, it shows you what kind of questions you can ask for the agent. It's going to start loading itself because it's a test environment. It seems like I have four overdue tasks. So that's good. Not good, not good, but good for a demo. So I could say, uh, those four. Move them out a week, you know, and I can actually just dictate to this as well.

Or I can type. So I think what you're showing too is not just knowledge, but it can also take action and help you do work, get work done. Right? Yeah. So it's human in the loop, right? So it always checks, right. And that's what we like about it. Because sometimes reps can kind of. Jump to conclusions and say, oh, you know, bulk up day 200 records. And so when it comes back, it says, hey, this is exactly what I'm doing. These are the changes that I'll be made. This is how long it'll take. Are you sure you want to do this? Right. Yeah. And because it's our, um, demo test environment, actually one of them failed. And that's actually a good to highlight that if it fails, usually the rep would have to run into an error message or a bug. The agent will actually retry a few times, try different methods and come back.

Oh, this might be a problem. You might want to reach out to the Pega team. Nice. Um, I probably just broke this agent because I was, uh, doing that, but, um, yeah, you can create and I'm going to go in the back end because I really want to show you guys the back end as well. So like. How many, how many, how many automations or tools are you building a month? Would you say at this point or a week. A couple a week? I would say. I can build an agent in about 30 minutes. OK. And a new tool. So if someone comes new tool. So agent. Tool. So if they come say they want to expand and I'll go to the back end here. Um, let's say that. So this is, this is the, um, the main definition of the agent. Oh, it's very small.

Uh, zoom in a couple of times for you. Maybe close the. There you go. It's fine. I'll do that. Um, and I'll show you how simple low code this is. I mean, I have a technical background. I'm a developer. I have, you know, I have all this knowledge, but it's literally just a, I could just do an agent instructions, which is extremely simple for my agent. I don't put, I like Claude in the planning tools in Pega, which they built in the back end of this. Be able to plan and understand the data it has. So I just kind of give it some response and style rules. These are actually links to our internal systems that explains how to send a link. So when the agent responds with a link to our internal system, it actually puts them in the right direction and they can log in.

And dive down to that. The system is breaking that down into an API call. Just just think about the power of that. You are telling the system. Reach out to this other internal system. Pass in this parameter in words. It knows how to transform that into an API call. Get the data back. Map it to the context and give you a coherent response. And that's in milliseconds. Yeah. So and I'll show you examples. So the org search, for instance, if I wanted to search for an org in our system. And this is a bit of a technical. So sorry for that, but I just wanted to kind of highlight how it works. Pega. I'm able to build a data page, which parameters. So essentially I design the parameters. So if we go back to the agent itself, you can see let me go to parameters we have created before date subtype, exact name, territory ID, you know, product if you want to go by product.

And the idea is that simple. I just have the agent passing these parameters. Build a query that then reaches into Pega and it's just parameters. The agent itself picks which parameters it wants to use, passes into the the agent on the back end and run the query. So you don't have to. Essentially, the days of running a report are gone. You could just have the agent say, I'm trying to do this, and it will figure out the parameters it really should be using and kind of. Talk to your data, use cases. Talk to the data. Everyone has that. And, um, you know, it can create random leads. It can, I think the best use cases are people can say, I, you know, if you even do, I met will at PegaWorld. Um, can you, you know, I want to follow up call with him next week.

He, you know, and set me up a meeting and build me an agenda for that so I can just you could just talk. Let's say you're talking at the event. You just I just met will I scan his business card? So yeah, usually I have like a I'll just do And Pega. I think this is a lot of sellers are getting super excited about. Yeah. Pull up your mobile device, talk to it. And so this would be more of a use case from a. Yeah. So what actually it's going to do first and this is important. It's going to look for will in our system and see if they're already there. So duplicates it's going to find duplicates. It's going to find their leader contact that's already involved with something. Um um is it. You can just say new one hit enter and it'll do that for you.

Yeah, yeah. Um, yeah. And I figured out Pegasystems like these weird things. Like if you just put abbreviations for companies, obviously. Yeah, it's going to start building some, some things. Uh. And so what's powerful about this is that it's stream of consciousness, right? You finish a meeting, you basically record a voice note and you put it in and then it says, hey, you know, here are some fields we need. We need the role. We need the company. Could the company be slightly different? You can you can connect it to LinkedIn and say, hey, look this person up on LinkedIn and get the data from there. Get it from Zoominfo, which is what we use. Yeah. And it does all of that for you. And so there you go. It proposes the lead and says, just confirm that everything looks good. Should I create this lead.

And then you can create it. It's going to create, but it's not only going to do that. It's going to do a few other things to ask it a few things. Um, obviously that's. A good point because that's your, you know, what are the steps that should be done? Right. And, and you're building that into your workflows. And that's why it's. So it's like, oh, I created him. I'm going to do a follow up task for you now. Yeah. But you don't like, and you can just write it like you could be talking your phone really fast. Okay. And here's the suggested meeting agenda. And then you can. Just write me an email. So it's, it's really just about those things that you have to. Yeah. It was great meeting you. PegaWorld. Did I even say I met him at Pega. I did say that.

Yeah, yeah. Okay. Just I was like, did AI just assume we're at my location? Um. This is great. So is your vision that sellers will just kind of live in this experience and they can just kind of talk to their assistant and. Yeah, because then I can say like, okay, this. I think just like creating ops, creating orgs, creating org hierarchies, like what product is it going to be? I can just say this mouse is not my mouse. So this is a demo mouse. I think that's the point, right? You know, it's the single pane. You don't need to go into the lead interface or the contact interface or the org interface. There's none of that. Oh, you forgot this field or this field is actually not permitted selected a role that's not valid. None of that happens. You just talk to it and it talks back to you.

I want to show that obviously to make an op, I need to convert the contact. I didn't even think about that. The agent's like, well, you need to convert contact. So I'm going to convert him to the contact. You want me to do all three steps at once? I think that's a great example. It's following the workflow that you've defined, right? It's not, I'm. Not making it up as it goes. Yeah. And you know, I didn't script this. I'm just you just come up with it and you can just keep on working with it. What I found is a lot of users then say, well, find me my top three leads that I want to talk to next. We had a recent marketing event, which ones are unowned and haven't been followed up with next yet. Find those for me, right? Build me a task list.

And then in a couple of days, when you go back to the agent, it will go, oh, you still have those tasks you need to do? Yeah. You need to follow up with those, those items and like, take the next steps. Um, and what this really opened the door is like, and then it also has the, this is just our base environment. And I'm not going to show you live because it has all of our customer information. But when I'm doing this kind of stuff in live, it has all of our context. It's just fantastic. You meet someone new, it's going to have all of our products. What opportunities are in the pipeline? Something related? Oh, we sold this way before. You should mention this. You should talk to this person. Bring a demo in. Here's the agenda for the demo. Do you want me to schedule that for you?

Because you can connect to 365 and say, oh, you're all set. Would you like me to have a competitor's med pick or path, which is a pricing tool doc that we have to do? So essentially the, the time it would take you to learn our process, you could just naturally interact with this agent, right? And understand how our business process works because it's embedded in the tool itself. Yeah. I love your excitement. I feel like. Yeah, I'm a very excited initiating from you. Yeah. When I went to like Germany and the UK and others where they have to do these crazy RFPs and like all this administrative work, they could just say like up, down update all from my recent activities, update all my opportunities. Are they in the right stage? Do they have the right percentage and just go through them that way? Then go back to selling the human interaction, right?

So this is a fact. I think a lot of people and are scared. Maybe I could just take all their notes now and just replace them. The idea is to enable them. Augment them, really. Augment them, but enable them cut down all that work they hated doing, and get back to the human interaction where they can actually. Like doing. What they like to do. That's great. And we're down to like, I knew this would happen. We have seven minutes left. So I do want to we should open it up to questions. Is there anything else you want to touch on the deck or. No, I think we can skip the last slide. I mean, the point to be made there really is just that. As Corey said, we went from zero to live in days. Can we move back to the PowerPoint? Yeah. Yeah. And to that point as well.

Zero to live. And I meant it before on the on my plane ride, I was able to build 2 or 3 tools on the plane with bad Wi-Fi. Like if you have a use case in mind and you want to start developing with Pega AI, you just need to build that foundation. And then all of a sudden it's just, this just opens up. You can, like I said, you can just say you want to have the org agent have a new parameter. I literally just go in and add a parameter on that low code page. Add it to the other data page in the query. If I want to write a note how to use it. You can, but you could be very minimal because it has the context of our data. So it understands what the fields and the properties. Are, and the link and the lingo is the same, right?

You're not reinventing the wheel here, right? It's just that your case management becomes an agent workflow, right. Your decision rules become decision logic. The flow design is orchestration. The data pages become context, and the guardrails are the boundaries that AI has. So it very seamlessly your existing infrastructure maps into the AI ecosystem very seamlessly. Yeah. Awesome. You kind of get consumed with solving use cases. I find it's, it's. It's a, it's addicting. Yes it is. I'm addicted. I'm working way too much, but I enjoy it very much. So I love working with Pega. So thanks for the demo. That was great. Yeah, yeah. Questions. If people want to jump up to the mic. If you're shy, we can meet after up the chat. Um, love to talk. I could talk for hours on this, so. Any questions? Andrew, do you have any questions if no one.

Um, yeah. I mean, one thing that I was wondering, I think you touched on this a little bit about prioritization, but and when you launch, after you launch a new feature, do you guys have a process today for capturing feedback live or how are you doing that? We use the, oh, you can't see it. If you put the demo back up for a second, there's the, um, thumbs up thumbs down here. So if you give the thumbs up and they can like or thumbs down, they can give some feedback. Also, what's good with Pega is I not a big brother, but I have we have all the conversations ever, right? Because Pega just natively does that for everything. So there's like a tracer, um, here with the AI tracer and the stuff in the back end. This is so powerful because. Yeah, you can see all the, all.

The questions that people have asked. Um, are people asking the same thing over and over again? Are they liking the answer or not? What tools are they using? So this is kind of your really gateway into seeing what people are doing. And yeah, you're able to actually see how many tokens. I mean even tokens are it's a low amount. It's irrelevant. It does a really good job actually. I think they probably spoke about it today, but the Pega use and be able to orchestrate and plan with the LMS. The token usage is. So look at it, 10,000 for all the stuff we did that I just did. It's like nothing. It's like nothing. Yeah. Absolutely nothing. And then here you can actually see. So if we, if someone goes, hey, it didn't really give the best feedback or it didn't give me a good result.

We can look at the actual data that the agent. So this is, this is the query I've shown you before, or one of them. It pulled the data in a Json and that's how it referenced it and used it in the response. So if you want to actually see what happened, you can see that what they asked how the agent planned to do those actions, the actual data. And then you could actually just feed all this right back into Cloud and say, what happened? So I could take this whole tracer and put it right into Claude and say, what went wrong? Oh, well, your parameters are off. You need to change this. It didn't have the right data. The context window was too big. This is all things that happen often. So what I did is, for instance, for a lot of data, I paginate it.

So if there's 100 results to the fix that context. I don't know if I told you that. Yeah, essentially we do 50 at once. And if it needs more data, the agent will make another call to reduce the context window and save tokens and also get a better response. Cool. Nice. So where do you go? So let's say you're here on this stage a year from now. What do you think you'll be showing or what's next for you guys? I think some of the things you guys have been talking about with like the prepare for meeting, it's more autonomous action. So just do your work in Pega. And then when you're about to have a meeting, it's already prepared. You. So essentially it's did all the prompts. It did everything for you so that when you go to the meeting, you can just now go to the chat window on that meeting and say, oh, you know, tell me about this person or tell me, and it's already there.

And I think some of the stuff where it's listening live, I think giving. Negotiations with teams, I think is one. Team's integration, just bringing it all together. And then just. The more context. You know, scaring people with autonomous action with that is kind of the next step. But hopefully, yes, please. Keep your data actually accurate. So maybe should, can you put the mic so they can hear you on the, on the recording? How do you keep your data accurate? I can imagine that there are a lot of contact persons in your CRM system, and one of the persons is going to another company or leaving the company. Is that a sales agent or also something. So third party tool we use for that. We use Zoominfo for that. And that actually sends us insights on any existing contacts, switching or switching to another organization. And then that automatically cascades through our CRM.

That, but there's a combination of a couple of things as MCP coming. Well, actually there is an MCP and there's more stuff. I don't know what Andrew can mention, but Zoominfo will be a part of Pega. I think we will ship. So yeah, that's a great question. And everyone struggles with contact data management, especially. I think what we're doing in Sales Automation is building a workflow that can identify and help resolve duplicate contacts or cleaning up contacts. And what's really great about these agents now is that's kind of hard to do if all you have is just the data in your system. But when now you can go to the web, do research, go to Zoominfo, do research. There's all the types of vendors. Zoominfo is great for US based contacts, but if you go to Europe, there's actually better vendors there. And so being able to connect all of those into a single agent in a workflow, we're starting to see that problem become a lot easier to solve.

So yeah, and even within the system, I mean, I was just, this is actually the new, we didn't have time to show it, but there's, this is like the new Infinity studio and the new Pega. But, um, it's okay, I think. I think I need to log in. But anyway, you can you can tie in all that stuff. You can manage your data a lot better. And you could even ask like, if there's a, you can you do it autonomously, or you can even ask the agent people lead management. For instance, if a new event happens, check out all the leads. It will actually look for duplicates. It will look for potential people that moved. It will go reach out to Tavoli or Zoominfo or others and aggregate the data and make choices to help you maintain the data. I think, yeah, I think the next big thing we should work on is using Blueprint to map out some of like that kind of workflow Other workflows that you guys have that are repeatable because then it's just easy for people to visualize this is what's happening and when.

But then the agent can execute all those things in a repeatable way. So that'd be great to tackle next. All right. Well, we're at time. We're at time that went fast. So thank you Corey and Nish that was great. Thank you Andrew. Thank you everyone.