Video: The Suite Customer Forum - Oct 20 | Duration: 3560s | Summary: The Suite Customer Forum - Oct 20 | Chapters: Welcome and Introduction (3.81s), Meeting Agenda Overview (109.91s), AI Confidence Problem (162.6s), AI Data Challenges (243.04s), Precisely Platform Evolution (335.65s), Unified Data Management (408.08s), Precisely Platform Overview (490.855s), AI Studio Demo (544.615s), Precisely AI Studio (618.08s), Property Analyzer Demo (711.17s), Property Analyzer Demo (822.61s), Address Validation Demo (1091.665s), Q&A and Transitions (1285.965s), MCP Architecture Overview (1397.775s), Live API Demo (1595.57s), Fleet Tracking Solution (1723.255s), Five-Step Process Demo (1962.745s), Results Review (2277.98s), Fleet Optimization Results (2334.21s), Live Demo Recap (2398.12s), Geofencing Q&A (2519.605s), Q&A and Transitions (2689.765s), Community Forums (2767.31s), Q&A and Adoption (2995.71s), Data Trust Challenges (3207.53s), Session Wrap-Up (3304.815s), Support and Next Steps (3368.07s), Closing Remarks (3479.87s)
Transcript for "The Suite Customer Forum - Oct 20":
Good morning and good afternoon to everybody. We would like to welcome you to today's version of the suite customer forum. We have a great program for you today, and we're really looking forward to, sharing some really exciting updates with you on the data integrity suite, Precisely platform, and, the Precisely AI Studio. As a reminder, we are using our Goldcast platform for today's presentation. I wanted to point out a couple of items for you. One, on the right hand side, you will see a chat button. You can use this chat button at any time to communicate with us, the presenters. Next to that, you will see a docs tab. You can click on that for some helpful links. And then finally, on... Next to that, you'll see the q and a button where you can press that and at any time ask a question to the presenters. We'll take as many as we can, toward the end of the program, and we're leaving plenty of time for questions. So make sure to get them in early and often. On the chat, if you would do us a giant favor, let us know where you're calling in from. You'll see that I just put in Fort Myers, Florida, where it is a beautiful 24 degrees Celsius, which is about 78 degrees here today in Florida. So it's a great day. At the bottom, you'll also see a thumbs up. Let us know how we're doing. And finally, there's a raise hand button that, when we get to the q and a, if you'd like to join us on stage and ask your question live, you are welcome to do that as well, and we'll review that again toward the end of the program. For our agenda today, we will have, two new members of our team, Joel Andrews and Rashid, come and talk about some of the changes that you may have seen recently, and, Rashid will give a demo of AI Studio. Then we'll have our friends from Quorum talk about some real real world use cases of the AI studio. Then we're gonna introduce some new user forums from my colleague, Matt. And then finally, Antonio, who many of you should know, will come on at the end and do some q and a. And with that, let's get started. Joelle, if you'd like to join me on stage, we'd like to get started, with our welcome. And, Joelle, they've come off mute, and it's all you. Please go ahead. You hearing me okay, David? You are great. Let's let's kick it off. Excellent. Okay. Let's get started. Alright. So have you ever asked your favorite AI a question and been very pleased with this confident answer? I'll make you answer this, David, since you're here. All of the time, Joel. It's always accurate. That is a very good answer because, yes. Me too. And, unfortunately, every once in a while, that, answer is not correct. And, yes, when we see that, we get really surprised and say, uh-oh. What happened? That's because the answer was wrong, and AI can trick you like that. Confidence is not the same as accuracy. Having a little technical difficulty. Don't worry. So that's the problem here that we're facing. AI doesn't stop to question when your data's wrong. It's gonna act anyway. AI agents make it possible to move from simply giving us answers to actually taking action. And there's four ways that we here at Precisely are seeing this show up. First, AI can't access the data it needs. A lot of critical enterprise data still lives in systems like mainframe and SAP that AI platforms can't easily reach. That means AI may be working with blind spots and not telling you what it missed. Second, AI trusts what it shouldn't. AI doesn't necessarily stop when something in the data looks unusual. It can act on bad data confidently, and then worse, it can scale it. Third, AI can't explain its decision. If you can't trace an AI decision back to the data that drove it, you don't have the transparency and the accountability that you need. And fourth, AI can be missing critical context. Your internal data only tells part of the story. Without enrichment and real world context, AI can still be making data... Making decisions from an incomplete picture. For years, organizations managed data through point solutions with different tools solving different data management needs. Those solutions worked, but as the number of systems, technologies, and point solutions grew, so did the complexity of managing everything together. In other words, bad data used to just mean a bad report, but now it can mean AI taking the wrong action and impacting your entire business. The stakes have changed. But you already knew that. You're here. When you chose the Data Integrity Suite, you recognized that managing data through a growing collection of point solutions wasn't going to be sustainable forever. In fact, according to one recent study, 70% of respondents were using at least five to seven different data management tools and spending about a third of their time making sure the tools worked together. But not you! You were already moving in a different direction towards data management capabilities that work together on a common foundation. You made a great choice. If you attended the Precisely Now event that was happening on October 8, you've already heard some of this. That's where we introduced the Precisely platform. The Precisely platform is a unified data management platform, and it's the next in the evolution of the data integrity suite. It builds on the capabilities and common foundation you already know while moving from a model of individually purchased services towards a more unified experience. The journey as a whole is about moving from point solutions to capabilities working together in the data integrity suite to a unified platform. The platform is currently in preview with general availability planned for the first half of twenty twenty seven, and pricing and packaging are still being finalized. But don't worry about anything. As a Data Integrity Suite customer, you are already on the right path. We're gonna back things up here just a moment and ask ourselves, what exactly is unified data management? Agents, applications, and analytics operate across the entire data estate. When every system and data management capability are managed separately, the gaps can become AI risk: fragmented data, inconsistent quality, and no shared governance. A unified data management platform creates a control plane across that environment. It brings data management capabilities, shared services, and enrichment together while allowing the data to remain where it lives. And that is how unified data management establishes greater control over the data feeding your AI. And so this is a look at the Precisely platform, and it's how that unified approach comes together. The platform brings precisely data management capabilities, shared platform services, and data enrichment together while connecting with a broader technology ecosystem. On the left hand side, you see the precisely AI Studio. That's where we're going to focus more today. AI Studio provides a simple way to start putting reliable data to work in AI powered solutions, and I am not gonna give away any other details right here other than to say you want to see this in action. So passing it back to David momentarily, but my friend Rashid, he's going to be showing you how it works. So let me give the reins back to David real quick. David? Thank you, Joel. Amanda, if you could launch the first poll for us, we do have another way you can give us some feedback. We'd like to know, if you are confident that your AI tools provide answers based on trusted and verifiable data. And we have three answers for you. Do you agree with that statement? Are you not sure? Or do you disagree with that? And we'll leave the poll up for a couple of minutes, and we'll come back, a little bit later on. Also, if you have any questions for Joelle, drop them in the q and a chat, and we'll take them back at the end of the session. Thank you, Joelle, very much. Alright. See you later. Rasheed, thank you very much for joining. Take it away. We'll get into a demo of the precisely AI studio. Thanks, David. I appreciate it. Can everybody hear me okay? You are great. Great. Awesome. Well, first off, thank you very much. I appreciate it. I just wanna welcome everybody. Good morning, good evening wherever you're joining us from. And I am personally very excited, for what I'm about to show you, which is our new Precisely AI studio. So jumping right in here, start sharing my screen. And just give me a moment. And if I can get a thumbs up in the chat or from my colleagues, you guys can see my screen. Great. Yep. Great. All set. Perfect. Awesome. So this is the new Precisely AI Studio. Jumping in really quick to why we built this. A lot of people, in various industries, in every facet of their work and their life, they've been asked to use AI and put AI at work, but they they frequently run into the same problem that a lot of other people run into, which is the data just isn't ready, and there's really no fast way to be able to find out if something is going to work with AI before you make a commitment to it. So as many people are aware, trying enterprise software, enterprise data and AI tools usually means, you know, sales calls, trials with heavy setups, you know, just general weeks of plumbing before you can really see whether it delivers. AI studio, precisely AI studio, is our answer to that problem. So what I'd like to show you here is a growing collection of ready made apps, agents, and skills that run on live precisely data. You can see it work on real data and decide for yourself, and then you can take any pieces that you want, use them in any AI model, that we offer here that I will walk you through, that you already have without really being locked into any, single stack. So let's jump in here. Where I'm sitting right now is a catalog of the various assets that we have. They're categorized by apps, agents, and skills. But what I find great about this personally is if you were to sign up, and we would love for you to, you're gonna get a curated sign up process that is going to ask you a series of questions about your interests, about what you do for work, and we're going to put you in a position to see things that are relevant to your industry or your interests. And over here, what I'm particularly excited about is, just the real estate stuff for me personally, and I will tell you guys why and they... Give you an example. I've already preloaded something that's called the property analyzer here. And, what this is here, just... Let's start with an app. So we have some pre made apps that you can jump into. Property analyzer, for instance, my personal favorite, and I'll tell you guys why in a moment. It's for anyone who needs the size of the property quickly, can enter any US address and click analyze and, you know, real estate intelligence agents, will run and, complete, the due diligence report using geocoding, address verification, risk scoring, property details, and, we'll deliver it automatically through, you know, precisely data APIs. But I'm gonna use for this one, pretty popular address that I have loaded in here already that you can come in and try in, property analyzer. And as you can see, without having to make way... Make everybody wait to load here, we're using 1600 Pennsylvania Ave, and we have a nice robust output of information, valuation information, tax jurisdiction, and most importantly, what I find very interesting is, nearby places or points of interest. But let's do something a little bit more fun, in my opinion, and I will load a particular interest that I am personally very interested in, which is 508 North Kaleheio Avenue in Hawaii. I think it's on the Big Island. I have never been to Hawaii, but I am, a stickler for sitting on Zillow and cranking up the price of a home all the way up to, like, a $100,000,000 and then, leaving comments on YouTube, videos about, you know, mega mansions of how this house must be sitting on a flood plain or my neighbors are too close to me. So when I first got into the AI studio, one of the things that interested me the most was, actually the property analyzer for that reason. So while this is loading, my dream home is loading here. Hopefully, it doesn't take too long. I will tell you a... An approved, dad joke just to pass the time. So it is... There we go. Now it's loading. Why do melons have to get married? David, do you know? I do not. But why do melons have to get married, Rashid? Because they can't elope, David. So I know. I think we timed it I think we timed it just right, David. So... Yeah. Without any without any practice either. That was awesome. know. Right? Oh, man. Okay. So as you can see, this is still loading, but what's gonna come up is, you know, various... It's property data here. So, there we go. Now as the record populates, we could see that there's a... At the top, we've got a composite risk score. We've got... Just under each of this is each single score that's structured. It's got flood risk, prime index, address quality, and where the data isn't available like, you know, earthquake data, for instance. It will it will let you know. So, but, yeah, this is going for 22 mil. It's actually listed at, I think, still 50 mil, but that's neither here nor there. I'm only 499,000,000... 999,000 some odd dollars short, but maybe I'll get there someday. So, moving on. The next thing I wanted to show everybody was an agent. And, again, just using the real estate, the real estate example here. Preloaded this. If you were to jump into the real estate agent, this can turn, US addresses into a unified property report, property building attributes, tax jurisdictions, again, the local time zone, nearby places. It's It's meant for investors and portfolio managers really who want property values, signals, property characteristics, tax tax implement... Implications. Excuse me. And it's built from our geocoding property intelligence checks, tax jurisdiction, time zone skills as well. And, there's some... If you wanted to hop in here, there are some just preloaded, prompts that you can just click on and try out here and see the, the outputs over here on the side. So for this one, I I personally... If I'm looking for a home like the one that I just showed you... The other thing that I'm interested in, how close are hospitals to this house if I, I don't know, stub my toe and I need to go get medical attention. I'd like to know how close something is just for me personally. So, now moving back this way, I wanna jump into something that I find really powerful, with the AI studio and I... That is the, connection to our precisely hosted MCB. And what I'm gonna do here is I'm gonna take on a role. I'm gonna use a sample prompt here that I put together. And this out. And, sample prompt reads this, here are three addresses from an insurance intake form. So if you're in insurance, somebody's giving you some incomplete addresses. For each one of these addresses, let's verify and standardize it and then tell me what's been corrected, whether it is deliverable, and let's flag anything that you can't confirm with confidence. And you know what? Let me go off script and make a tweak here really quick. Let me or help edit In the table. And then always, I believe it is good practice to thank the AI tool just in case. Alright. Let's run that really quick, and let's see what we grab. Excuse me. I don't have a second badge yet, David. Well, I was just gonna say I always thank my AI agents too. Just, you know... I I don't know why, but I'd say please and thank. you. I do. anthropomorphize it a lot. Yeah. Yeah. Me too. I I I call it dude. I ask it how it's doing. Alright. So here we go. So here is a nice, output table here. Some of the addresses that I've given in the same prompt. As you can see, connected to the precisely hosted MCP here, we're actually able to just query, directly from the MCP and, get get some really good information on these addresses and some fixes on the addresses if we wanted to as well. So first address here, we can get, you know, it is deliverable. Yes. But it's not flagged as, you know, non deliverable because of the ZIP code. It's got a score of 80. We can get a whole bunch more information better. We make some better prompts up here, which I can do in just a second. We've got better follow on prompt, actually. So let's do this. Again, so, again, for our insurance, example here, for the addresses we've confirmed, the the... Claude has been nice enough to help us with confirming. Let's pull the property records. We can use a year, build, size, stories, market value. And, again, something that I'm always concerned with personally, FEMA flood zone, with, with, the elevation. Let's go here. Give this a shot. Again, I made sure to thank my AI. Alright. It is loading. And while that is thinking, David, I do wanna say, up here, yeah. So here's a nice table built out. So we've got use, can be commercial, could be residential. At least I'm not gonna do anything residential past my dream home. Gives you the year built, size of the, square footage, stories, etcetera, you can see. So, really, what I just wanna highlight here is just, use my own personal, interest in guilty pleasure with, maybe unhealthy guilty pleasure with how much time I'd like to sit on Zillow looking at homes. This is just something personally, interesting to me, but we are joined today with a, a valued partner, who is going to be... In Quorum that is going to be able to, show something quite a bit cooler than what I have shown you, and they did it all using the AI studio. And I'm very, very proud to, hand you back to David, who will, Rashid, tee it up. actually have a, question for. you. Chris would like to know. Can can you show us the initial prompt again? Oh, well, for which one? Sorry. For which one, Chris, if you could answer it in the chat there? or the first think? in the second one. Yeah. And, also in the chat, Antonio put a link, the first. one. Sorry, Rashid. If you could go back to the first prompt. Interesting. Oh, first prop. Excuse me. Okay. In the chat, thing. Antonio put a link to the Precisely AI Studio, but also at the top of your screen, you'll see a purple button that says Precisely AI Studio. If you click on that. button, it will take you right there for the free trial. Oh, okay. Somebody asked... Excuse me, David. Somebody asked a question to see the first prompt. Yep. Is that correct? The first. prompt. Okay. There you go, trying. to find the questions over here for people. And I can also answer your questions if if I come off stage. too and and interact with this, And, whoever, has questions. while, we, sir. why don't you, copy and paste that prompt and put it in the chat as well? That would be, Okay. handy for Chris. Let me put my thanks in there as well. Thank you very much. And with that, Rashid will be back for q and a later. Yes. Thank you very much, everyone. Appreciate it. Great job, Rashid. Thank you so much. So let me share the slides again, and we're gonna go to this one. And I'd like to welcome to the stage, Nora and Scott from Corem. We're gonna talk about some real world case studies. They're a great partner of precisely based in Canada. So, Scott, Nora, take it away. Thank you very much, everybody. Get the first slide going here. Alrighty. So thank you, everyone. I'm Scott De Deceno. This is Nora Tremblay. We both work for Carem. We're a Precisely Partner. Today, we're gonna start with a a little, I guess, you'd say MCP one zero one. This was, actually designed for a previous webinar that got moved back. And then the AI studio came out, and Nora attacked it very well. And then she put together a sample application that was a real world sample, customer engagement we were working on at the time. Alrighty. So I kinda broke this down whenever I was starting to learn this stuff about kind of an old way to do stuff and a new way to do stuff. So the old way, we would have... Everything is spelled out. Everything is created by developer. You'd have an address input, and then you'd click it, and then certain things would come up. So you can see on the left side here sort of a... An example of an old way that things were done. Whenever you had to spell out everything, if you wanted to add anything to it, then you had to reengage the developers and reengage the the process, and it took some time. As opposed to with the AI and the MCP, the MCP, when you learn about it, they say it's metadata, it's metadata, and that drives everything. And so now with the MCP, you're able to just ask an open ended question or ask a question. And you can see a prompt on the right side here of what the question was and how we get the information back. Now one of the positives about this is that you do not have to reengage developers. You ask a different question, of course, presuming that everything is connected on the back end. So you can see the architecture of the two different versions. On the top, the old way. On the bottom, the new way. You can see on the right side is the exact same thing. So this goes to, in this case, that I'm gonna show you, it could go to what we call the graph API. It's an API that touches on property datasets. The front part of it, where you used to have a developer create the front end, now we use the LLM. The LLM, the large language model, that kind of understands that English language, understands what's going to be said, etcetera, and that all drives off of the MCP. So when we do an MCP, this top left part here is kind of some of what shows in there where you go ahead and you explain your fields. On the bottom, I have some field names and descriptions. So this original MCP that we used was more open source. It wasn't maybe as developed as what they have now in the AI studio. And when we looked at it, we know we can take... So the the description and the field name, that's out of the documentation for the property datasets, and we could enhance this MCP that was open source and add more of that descriptive stuff that, of course, the LLM can understand. So I kind of, in my head, kind of put this as a USB connector. So you know how we can go to all these different entities, you know, your printers, keyboards, everything else that used to be different connectors. Now we go ahead and we have the USB to connect everything. So it's kind of that universal idea, that universal way to connect or share information, and it's built off of, as we said, the the metadata. The metadata is this long. descriptive stuff. It's information about information, and that helps the large language because when we see this description, if I just see doc type, I may not know what that is. But when I read the description, understanding the English, then I'm converting that to something I understand. Similar idea to what's going on in the back end with the AI. So I'm gonna jump over to a quick little demo here. I gotta share my screen. Alright. You guys should be able to see precisely Datagraph API demo. And what we're gonna do is we're gonna just gonna, similar to the last one we saw, everything... A lot of times in the geography world, things run off of an address. So we're gonna put an address in. And then you see this idea of the old way to do it. So we have these various different tabs that might tell us information. I'm gonna just click on the property, and it would give us information about that property. Now all this stuff had to be developed by a developer, etcetera. Now when we go to the AI, all we gonna do is we're just gonna put a question in here, same question you saw before. So this should look somewhat familiar. And we asked that question, and now it's going through the MCP going back to those same APIs that we did off of the, developer created stuff, and it's asking that question. And we didn't do anything to speed things up. And so now it comes back with all the information that the AI or the MCP said was, germane, etcetera, to it. And I can just ask a different question if I want to look at something else. So I'm going to stop sharing. You need to make sure we move along here fairly quickly. And then you guys should be back to the slides. Is that correct? Alright. Yep. So the idea, of this whole thing, it's scope for a purpose. It answers what the tool covers, and it doesn't answer anything anymore or less of it. And you're going broader is a choice. So where the line is, if you ask a very, very broad question, in on the AI side, maybe that means you've got to do a lot of transactions. Sometimes those things, that's where the... I wouldn't say the trap is, but that's where the line is. So your questions, if they're too broad, may not give the answer back. So that's... You might say the benefit of the old style is you only ask one thing, but the the restriction is you can't get more information. However, you gotta understand on the question side of stuff how to answer a question right. That might be why the person said what was the prompt you asked because sometimes the skill is learning how to do that bit of a prompt in there. Okay. The Data Love tour was built before the AO studio existed. The AO studio... AI studio is the answer to the last question, how to broaden the purpose. You package your tools to work. Three things it gave us. One, precisely MCP server with about a 138 DIS actions. So we wrote no API integration code, a standard package, skills, agents, and apps that were published in one catalog, and the same skills also run outside of AO studio. For example, CLOT, when AI studio came out, Noor took it to a real opportunity. We're working on, when we work... And she did in about two weeks. That's the second part. So... Sorry. I have to move this way because my notes are not coming up correctly. So here's the opportunity. It was a US fleet owner. They had trucks they would take to repair shops. So the idea when they took these trucks to repair shop, they wanted to know when it was at the repair shop and when it left the repair shop. They they just had x and y coordinates. So one of the things that we do have is we do have all the... Or they would have all the repair shops that are okayed to be there or to use. So some of the questions that they might ask themselves is how much road time do we lose to repairs, which shop or which vehicle is the problem, and where should we send our vehicles instead if there is a shop that is a problem. On the left, we have what the customer gives us, the GPS points, the shop address. Everything else comes from Precisor. For the demo, it's a simulated fleet that Nord generated 24 vehicles, eight shops, six months of data. On the right, precisely data that does the real work. Parcels give us the exact outline of the shop. We might call that a geofence. The places have about 38,000,000 of interest. So if we wanna find a different shop, this is where we might go. And routing gives us drive times to the depot if we need that. And geo addressing, of course, turns that address into coordinates. And all four are coming through the MCP server from Precisely. The main idea, a truck arrives at a shop when its GPS points fall inside the shop's parcel. It leaves when this... When when they stop being inside. So why the parcel and not a circle? Circle... The parcel is more exact. The circle could get other properties. Also, the property could be bigger than a circle. How do you come up with a circle that defines all properties? It's kinda difficult. There there is... Time and time again in the discussion with the customer and what they tried in the past, there were more caveats where it was a problem, so it just came down to the parcel was just better. Two simple rules rules, a gap of more than sixty minutes starts a new visit and stops under thirty minutes are ignored. Nora will now show you how it all fits together. Nora. Thank you, Scott. So, yes, I'll advance the slide. So, error is how it works. So it's, the process is composed of five steps. The filled blue circles are caused to precisely, and the outlines one are our own code. Those don't do any caused to precisely. So the first step is the the the... Where we find the shop's location. We only need this when the address is not clear. The second step is a special search on precisely parcel data, and it gives us the outline of the shop slot. That outlines becomes our geofence. The third step is a... It's our own custom code that we've built on a... It's on a Python. And, the... It's... It takes the GPS points and checks which ones fall inside each parcel. That's how we could detect the visits. The next step, number four, it, where we scores every shops and every vehicle. We calculate the the median time per visit, the long tail, and the repeat visits. And the last steps, it's... We're then go back to precisely. And there, we find the report shops near the worst one and then rank them by drive time from the depot. So you may ask, who picks those precisely actions? It's the EIA through MCP. So it's out for... Of about the 138, tools that are made available through the MCP server. The the the AI picks the right ones. We've built skills that that tells which one works, and it's the same idea that Scott showed earlier. The the answer is in the description. So I'll show you a live demo to to it. Let me share. Thank you. No. Not this one. So as you can see, we took a lot of inspiration of the AI Studio look and feel. So we have built skills and agent, and let me show you what the agent look like. So it's... Or at the download... Downtime control tower. So it uses the skills, the agent, and the precisely MCP server. You can see the architecture here. So the agent will cause the skills, and each of them calls the precisely API platform. So let's launch it. So we have a prebuilt data here. So I'll run the analysis for all of them. You see here it's doing its call for to find the nearest repair shop. Sorry if it's longer that I would have loved to. See. Keep calling the MCP server for more data. While it runs, maybe I could look. We already have those results about the shop performance, the vehicle availability, and the calls are... There are our results. So we did the shop performance, the vehicle performance, suggestion for the best alternate shops, and a pattern. I... As you can see here, there is something. Yeah. So that's it. I think, the trends. We have the trends for by month. So let me go back to you, Sure. Scott. And... Yeah. Alright. So please note, like as Nora said, it's sim... Simulated fleet. The agent found all 101 shops from about 270,000 GPS points. No misses and no falls. The worst shop, Lakeshore, has a median visits of a hundred and one hours. The fleet median is forty seven hours, so more than double. Over 38 visits, so it's a real problem and not bad luck. The agent also noticed the trucks dropped off on dropped off on Monday stay 1.7 times longer at the shop, a likely reason the parts are ordered on Monday instead of being in stock. Where to send them instead? The agents picked Fullerton Service Group. The fleet already uses it. It has the best record, 10.8 median, and it's eleven minutes from the depot. Moving half of Lakeshore's visits there would recover 1,700 truck hours over the same six minutes... Months. Next, Nora will show you the same skills via, outside AI studio. Nora? Yes. As Scott told you earlier that there... We can, can see that we we have one server and many clients. So this one... This client was my cloud code session in Versus code. The the the live demo curse, once again, strikes. So I can show you on my laptop. And... But by by the way, I I took maybe a bit more time than than expected for my previous presentation. So as you can see here, directly in cloudcomm, I can... We, installed everything I showed you earlier in the app as a plug in in cloud. So right through the prompt, I can ask the same question, find a repair shop parcel at this address, and the same DIS calls are made. And I got I got the same answer through cloud code. It could have been cloud desktop too. So it's the magic of plug ins, skills, and apps. So I get you, Scott. Yep. Thanks, Nora. So three things to say, Cole. One, anyone can ask, the same precisely DIS APIs you already have, and now people just ask in plain language through the MCP. Two, good descriptions matter most. They are what makes the AI pick the right data. Three, AI studio packages, what works? We took a real fleet opportunity to a working proof of concept in about two weeks. So how long are your trucks not on the road? Now the fleet owner has answer from GPS points they already have and precisely data. If there's any questions, we are here. Thank you very much. Great. Thanks, Scott. Thanks, Nora. Really interesting, application of the technology. It... Great job and great demo. We actually did have, two quick questions for you, and and it was around the the building and the geofencing. The the first is, does Precisely do the geofencing? No. We we use the the... In all in all situations, we're using the various different functions from DI Suite to get the Polygon. And then that can happen. The customer, they don't need to ask and and take... You know, a GPS might get you back a point every three seconds, one minute, two minutes, whatever, and you don't need to go back, and there's not a point in Polygon in the DI Suite right now. So we're getting the Polygon from the DI Suite, and then you can store that polygon in your own system and you have your own GPS coordinates and you can do your own point in polygon that way. So no, we're not doing the point in polygon inside the D I suite. Terrific. And I I think this is sort of related, but, when you were showing the map, why not just draw a circle around the shop address? Would that make sense or no? Or Yeah. So, we mentioned how the... They wanna know when you're in in the... That the polygons can be a different size than the circle. There are triggers that happen that they want to know when it's in, and because those parcels can be much different. So some of the places, like, I take my vehicle to a a Firestone, and that Firestone is in a strip mall. And so the parking lot that I could put my car in or they could put the car, and they parked it different places when they're working on it, etcetera, it can be much broader than what that small circle would be, and they need to know when it's on the property, when it goes off the property for various different things. It's less of a security it's been stolen and more of a are they working... Is it on their lot? Because if if a vehicle needs to be be fixed, what they do is they take it to lot, and there's a trigger once they get on the lot. So a circle, if you said, give me a circle of 500 feet, well, that wouldn't work for Firestone or many of these other places. So they wanna know when it's on property. It triggers, hey. It should be worked on. And if nothing happens, then they can follow-up with the shop, where is this vehicle at for the fleet, etcetera. So the circle idea is something that they've tried before, but that really came down to the really positive for them was having that parcel. And because we also have, there can be on a strip mall or something like that, multiple parcels on that strip mall, We also can use various functions in DI Suite and with the property where we can find the common owner and combine multiple properties. So if that's an issue, we have that fallback too. So, again, it's a base where the circle ID breaks down, and there's other nuances that go into there that we benefit from the property data and the property, attributes that are with it. Yeah. David? Terrific. Terrific. Thanks very much. We'll see if any other questions come in for you. What what I like about this application is that if you expand it to other industries, right, it can be any asset you can track, right, in in any sort of environment or location or anything like that. So great. job. It was very impressive. We actually have another poll since we were talking about MCP. Amanda, if you could launch the second poll, and we'll keep that up for a a couple of moments. Our second poll, thank you very much, is are you incorporating MCP servers into your daily workflows with precisely or any other data tools? And, the answers are we're testing it, we're implemented it, or no plans to use, you need to know more, or you're not sure. And we'll leave that poll up for a minute or two. Thank you, Scott. Thank you, Nora. We'll see you during the questions and answers. At this time, I'd like to introduce, my colleague, Matt Malcheski. Matt, if you could join me on stage, and Matt's gonna talk about the precisely user forms. So, Matt, go ahead. Great. Thanks. Great to be here. Thank you for having me, David. If I can just share my screen, I will be able to pull some stuff up and show you it live rather than walking through some slide where I'm sure everybody appreciates being able to see things live and and direct, so to speak. So what you should be able to see here, is what used to be called the knowledge communities. It's recently been rebranded as the user forums, and as you can see, the interface looks a lot more polished. We have some nice contextual drop downs at the top, which are going to allow you to get to a variety of different support options, which make life easier for you. If you have not become a member, when you come to community.precisely.com, I'll throw that link in the chat for you all as well so you'll have it handy. You simply click on become a member, and it will take you here to just a brief registration portal. Now why you might want to do that is because, being registered allows you to post questions within the community itself. You don't have to be a member in order to view the communities and to look at the answers and look at the articles. But if you had a question you wanted to ask, then certainly you would wanna be registered. It's just another layer of interactivity, which I think you'll find beneficial. Also, if you happen to be somebody who is submitting, cases or support tickets, you will certainly need to be registered in order to do that as well. So going back to the the main interface, if I were to look in all forums here and I am looking for the suite forum, it simply starts typing data integrity suite, and there we go. It will contextually pull it up. And within here, you will see that we have a a good sized community, with a number of, individuals within it. This being the product community. Right? So the product community is the first one I wanted to show you. Within here, you're going to see, information coming out about upgrades, patches, versions, things of that nature. So in order to stay up to date with the current updates for the product and the solution itself, you'll certainly want to have this one as one of the two communities to to visit. There's also a discussion group, and I'll go back and and show you what that one looks like. Let's just go the the long way. I guess my machine needs its coffee this morning as well. So if you go into the, the discussion area, you'll see that we have about 50 folks in here right now. And what this area is designed to do is allow you to interact with we, at Precisely, our subject matter experts, as well as each other. So you'll see that there are a number of conversations which are ongoing within here. We have things around, ironically, the model context protocol, which we've heard a lot about today. Obviously, we have, announcements about upcoming events, some feedback from users we always like to take into account, and we have quite a few posts around, you know, how to leverage and get the most out of your suite experience. So, this is a great place to go in order to further your knowledge, connect with other experts, connect with others in terms of asking questions and potentially giving answers. I should mention that we have quite a lot of content in development as we speak. I don't want to spoil the surprise as it starts rolling towards completion, but I'd suffice it for me to say that there is a lot of content which is going to be starting to come in here that you will really only find in this space. So I would recommend strongly going to community.precisely.com, looking for the data integrity suite groups, both the product announcement and the discussion areas, and starting to further your knowledge and and get the most out of your investment within the products. So with that, David, I can pass it back to you. Unless there's any questions, I will, make sure I throw the link in the chat. And, folks, you can get to the the the the sweet discussion forum through the main, main interface that I just... That I will be sharing in a moment. Great. Thanks, Matt. And, yes, it is easy to sign up and it is free. And also in the doc section of our, space today in the, I actually did put the links directly to the data integrity suite, discussion forum and to the AI studio discussion forum, which is brand new, and we look forward to seeing you there. And Antonio, I see, has already joined me on stage. Thanks, Antonio, very much, Of course. for a little bit of q and a. So we actually had a question come in from Will, a little bit ago. For customers already using the data integrity suite, what would be the recommended first step for adopting the AI Studio and MCP, capabilities? Absolutely. So I'd... And I had mentioned this in the in the chat a little bit above, but what we are do... Recommending is that for... Even for existing customers, that you are signing up for a a free trial of AI Studio. So if you just go to precisely.ai, you could sign up for, for a free trial, there. You will need to, within the data integrity suite, change your default workspace, to the trial workspace to use it with, the, the AI studio so that you're not, pulling from all of the the credits and transactions, your existing, entitlements that you that you have. You wanna make sure that you're pulling directly from, the the trial space. And that's gonna make sure that all of the different AI studio assets work as intended and that, and that you're not pulling from from your existing, entitlements. So that's... That would be the the first thing, just talking about AI studio. There's there's roughly 25 assets, in there, right now. It's gonna continue to grow as Joel and Rashid had had mentioned, before, and there'll be new assets added, for any number of different industries and, use cases, that you'll you'll be able to test. When you do sign up for your free trial of the AI studio, it gives you the opportunity to connect the MCP server to your, let's say, your your cloud instance. I know that was one of the examples that was provided today. But there are other, models that you can obviously connect, the AI studio to, as well. So you can go through that process of of signing up for the free trial, making sure that you're using the the the right workspace, and then connecting the MCP, to your, you know, to your preferred, model of choice. So that that would be the the various ways to go about it. And that is with the AI studio that is, a hosted MCP, so you don't have to to worry about downloading and deploying that, yourself. Great. Thanks, Antonio. I think there might be other couple of, questions in there. So if, Joel and Rashid can join us back on stage for any additional questions. I did... While that happens, and we queue up another one. Wanted to circle. back on the polls. For the first one, the confidence in the AI tools, 80% of the respondents were not sure, and then 20% disagreed. So the the question about are you confident your tools are providing answers based on trusted and verifiable data? So that's what we're trying to solve for. And then, It's very interesting. yep, and then. What. do you guys make of that? I don't know. Antonio, Can you say do you make of that? can you say that one more time? So 80% of the AI people that responded to the AI tools. say they're not sure that the answers are based on trusted and verifiable data. I think that that's pretty, consistent with what we're seeing in the marketplace. I mean, there are depending on... You know, you can pull stats from all over the place, but, there was one recently that I thought was really interesting from, one of our other partners, at at Dun and Bradstreet that had done a recent, study on this, and they were getting, sort of interesting figures back, from their respondents that were consistent with what we were seeing a year or two years ago, on the same topic. And that's somewhere between 95, that kinda 93 to 95% of folks not necessarily trusting the underlying, data that, that they're, you know, putting into their their AI and ultimately seeing those results on the other side once you're getting some sort of output. So I think that's pretty consistent with with what a lot of people are experiencing right now. Definitely not alone. Interesting. No. Absolutely not. And then the second one on the MCP server, a little bit of shift from the last time we asked a similar question. Are you incorporating MCP servers? 60% are testing, 20% implemented, and 20 no plans to use, which last time we asked it, there was a lot of, I don't know what MCP is, and. we're not sure. what we're using. So there's definitely been a shift. That makes a lot of sense. What was the the percentages again? Can you say that one more time? 60% testing, 20% implemented, and 20% no plans to use. Well, that... That's fantastic that there's so many people giving it a try. Great to Great. hear. And if you're... And if you are using, APIs already, I mean, it's a great way to, offer another avenue for interacting with those with those APIs and being able to to call them, through, you know, plain language. Obviously, you may have those embedded into a workflow that isn't isn't requiring you to to to do much, but, it is a new way to be able to interact with them and solve some some interesting challenges that you may have. Great. I did actually just look at the clock, and we're running up to the top of the hour. We might have time for one more question. Antonio, did you see anything interesting. that came in? Yeah. One was, can I invite my team to my AI studio, workspace? And that's that's a great question because, obviously, you know, you have... You signed up for an AI studio trial, but you also want to be able to collaborate with your team on on perhaps a project, a data and AI project, and want to use the same AI studio assets to do so. So inside of the, the account tab within AI studio, there is a team section, and you can add members of your team. The, and what that's gonna do is it'll add them to that, that workspace, and then you'll all be working against, the same allotments, for the thirty days of the of the free trial. After that, if you are interested in in, in going further, obviously, there's the full, data integrity suite, deployments that, obviously, with you being... You all being customers, you have access to the to the data integrity, suite. And so you would be able to, use the AI studio, assets, or maybe you want to pull those down, compose them, assemble them in new and interesting ways, or change some of the skills, themselves so that the instructions are slightly differently or or work a little bit differently. And you can use that against your, existing, workspace, again, ensuring that you have the the proper entitlements that are... That you're calling on those various assets. But that's how you would... That would be the natural process would be, you know, working through your AI studio on trial, seeing if if that... If those assets work for you. And then once, you've established that, taking them, down and and using them with your production workspace. Great. And I assume if for questions, customers can... Should reach out to their account executive or CSM working on their implementation, and then we'll get the right team, in contact with them through those through those channels. And then the other the other place that I'd offer, those are absolutely... Use those those two channels, but the other one is what Matt had mentioned before. We have we have, been answering some questions from, from users within the, the, the community as well within the the user forums. So if you do have some, some some quick questions that you wanna ask of how this AI asset works or, you know, is this is this what I'm supposed to be seeing? Those are great... That's a great place. But, yes, of course, work with your account executive and your. CSM, as often as possible. Great. And we are running up to the top of the hour. So I'll just wrap up real quick saying thank you very much on behalf of all of Precisely for your time today and joining us. We can't do what we do without customers willing and, able and interested in adopting these solutions. So thank you very much for joining. We will be in touch with, follow-up materials. The recordings will be available to share with your teams, and we look forward to hosting you again on our next, sweet customer forum, or whatever iteration that takes in the new year. So thank you very much for your time. Have a great afternoon, and thank you for joining us today. Have a great day. Thank you. Thank you all.