Name: Contract Intelligence Platform Tour: Use AI to Automate Audits and Compliance
Uploaded: 2025-06-26T18:16:04.518Z
Duration: 1 h 12 s
Description: Contract Intelligence Platform Tour: Use AI to Automate Audits and Compliance

Transcript for "Contract Intelligence Platform Tour: Use AI to Automate Audits and Compliance": Yes. Compliance is quite ubiquitous, and there's a lot of places we could go. So for folks beginning to come in, we're gonna give folks a minute or so to get in. I see a lot of folks joining now. We're really excited for today's conversation as we'll be diving into how you can use AI to really support your organization's needs around compliance and audits. We're both gonna be talking to you and showing you slides, but there's also gonna be a demo. And as always, it's gonna be an interactive session. So please do feel free to ask questions as we're going through. Like I said, we've given folks a few moments just to jump in as we, look to start today's session. I think we've got a really good amount of folks now. I think we'll just go ahead and start going. We're really excited to dive into June's contract intelligence platform tour as we're gonna be discussing how to use AI to automate audits and compliance. Do wanna let everyone know, today's session is recorded, and so you'll be able to watch it on demand after the webinar has concluded. Just go to Workday's website, ZAFT, for words for that. We even give you a QR code at the end of the day session on how to access the on demand content. And there's also q and a. So please use the q and a icon here to ask questions as they arise during today's session. We have folks who'll be answering them. We'll also do our best to answer some of them during our conversation as well. And lastly, I do wanna share this product statement. You know, to the extent that we talk about any unreleased services and features, please do keep these statements in mind. Excellent. And with all that said, I'm really excited to introduce myself and Lauren Guidas. My name is Mene. I'm Wudiwe. I'm part of the founding team of Eversort, which is now Workday contract management. I'm an AI evangelist here at Workday. And I'm supported today with Lauren Guidas, who's a senior solutions consultant and is gonna be diving into an exciting demo. Lauren, excited to have you alongside me today. Yeah. Great to meet you all. Lauren Guida is here. I've been with Evisort for about three years pre Workday acquisition. Navi's gonna start us off with the, slide presentation about how Evisort can help with your audit audit and compliance, needs. And then I'll be running us through the platform itself at the end of his presentation so you'll actually get to see the product in action. So thanks for having me. Awesome. Excellent. I know folks are really excited to actually see the platform and technology in action. Excellent. Now what we really want to focus in today on is your contracts, your business documents, your invoices, etcetera. Because when we think about audits and compliance, a lot of those audits are looking for information that is somewhere within the business documents that you've done either within your organization or with other organization. You know, we really see contracts as the lifeblood of a business. And when we think of them as the lifeblood of the business, especially in today's world, we don't just see them as bespoke legal documents. We really see them more as data. And so much of our company's approach to contracts is born out of this view. The contracts are really data. Every business relationship, if you think about vendor relationships, customer relationships, even employee relationships that an organization has and is scrutinized, all has an associated document contract, invoice, statement of work, etcetera, with that relationship. And so if you approach those documents as structured data as opposed to the unstructured data that they are as they are just kind of PDFs and DOCX files floating around different repositories, you will find a load of valuable information and insight that's gonna enable your organization to be better prepared to deal with regular audits and also maybe even unforeseen compliance risks and issues that may arise. Okay? And this approach of approaching contracts as data, that's really been within our organization's ethos even before our acquisition by Workday last year. I'm part of the founding team of Eversort. We actually were law students at Harvard, data scientists from MIT, and built this out of the Harvard Innovation Lab with the idea of, let's structure the unstructured data of all the important business documents organizations have so that when issues arise, it's not a fire drill. We are going out to outside parties where they have to manually review lots of documents, to get answers on a on a particular timeline, but instead, you're able to just query and ask questions to already structured data. And this is enabled even more greatly by the generative AI capabilities we've brought on. We can just ask natural language questions to your documents. Right? But this approach of looking at your contracts and your important business documents as data and not just as documents and PDF files is something that's really gonna underlie today's conversation. Excellent. And when we think about these kinds of documents, I do wanna be clear because I know we've got folks from all over the organization today on today's call. You know, this affects folks in the office of the CFO. We're dealing with customer agreements, supplier agreements, trying to manage, you know, renewals, rate increases, rebates. It has effect for folks doing compliance reviews in the HR office. Right? I think even just recently, there's a a big change from the federal perspective of how non competes are approached. And now a lot of non competes that are in employment agreements need to be addressed by various types of organizations. Right? Just one example of the types of document reviews that could be more efficient by leveraging these kinds of technologies. And, of course, the CIO, the CXO office, anyone dealing with partners, technology vendors, looking through at their different data breach notifications or other kinds of confidentiality provisions, These are issues that, although legal is often tasked with managing contracts because they're written in legalese and in terms that lawyers understand well, the issues that stem from the management of information in those documents are important to the whole organization and manifest as issues in completely different kinds of ways. And so in talking about how we approach these kinds of issues from a Workday perspective, I do want to take a few moments just to talk about the two products Workday offers to actually help you manage these contracts and other kinds of business documents. One is our con Workday contract intelligence, and the other is Workday contract life cycle management. Now if you see here, the first step is, of course, connecting with where your documents are. You might have contracts all in Salesforce. Maybe you have repositories like Box or Google Drive or SharePoint. Right? Regardless, before you can get any insights out of them, you're gonna need to bring them into a system. Now what's very unique about our capabilities is that we can actually sync with these contract repositories in a very simple way where you don't have to disrupt the day to day uploading of documents, maybe to Google Drive or Box or SharePoint, but you can now get access to all of that. Because once we connect with these repositories, we then analyze the documents within them using advanced artificial intelligence. So even if you just know that that folder maybe has some agreements and some leases and maybe some statements of work within we can analyze 200,000 documents in under twenty four hours. You'll quickly be able to know, actually, 30% of those documents were leases and the wet the rest were maybe customer agreements. Right? But this ability to really investigate and turn those scanned documents and those folders of different kinds of agreements into structured data, pie charts, bar graphs that are dynamic and that give you visibility into the underlying information that actually has business impact for you. That's the important part of this analysis. And you'll see here that these two capabilities connect and analyze, this is all within Workday contract intelligence. Think of Workday contract intelligence as a tool that gives you visibility into your existing landscape and ever growing landscape of key business documents and contracts. Now Workday contract life cycle management encompasses all of these capabilities within Workday contract intelligence. But in addition, you have acceleration capabilities, which actually allow you to draft, negotiate, and sign new agreements even faster using AI that helps you actually draft these agreements, negotiate them, and using AI to redline them. Now for the purposes of today's conversations, really looking at how we're gonna streamline compliance and audits, we're gonna spend a lot of focus on this connect and analyze piece because a lot of addressing your audit and compliance concerns is gonna deal with understanding information that's across your existing landscape of documents, contracts, and obligations. But I did want you to understand the difference between these two SKUs in case they're brought up in other ways. Lauren, anything to add on this point? I think you mostly covered it. When we get to the demo, as as Memme said, I'm going to focus mostly on the connect and analyze piece with a heavy emphasis on the analysis side. But just know that it is an end to end platform, meaning all the AI that we leverage to turn your contracts into structured data points is gonna be very relevant for your new contracts that you're generating and sending through negotiations as well. So that's why we've tied it all into one singular platform. Excellent. Yeah. And, you know, moving forward, when we think about contract intelligence or compliance, you can even see a quote here from one of our customers, Western Union, talking about how, you know, Eversource capabilities being instrumental to their global procurement and legal operations at Western Union. I think what's really important is just thinking about bringing the visibility of the risks in your documents to into, sorry, into these kinds of dashboards where you can actually visualize risk. You can see here an insurance obligation tracker, where folks are using tools like this to track certificates of insurance and other kinds of obligations within insurance, be it making sure that vendors that maybe have legal requirements to upkeep certain insurance obligations actually have those in mind, and then also making sure that you internally have the kinds of insurance obligations that you've promised other candidate parties. I did want to take a few moments, though, because insurance, I'm sorry, because compliance is such a broad term. You can think of compliance from the perspective of government regulations that you need to deal with. You can think about the regular audits that maybe some regulated industries have to go through over time. You can think of things such as, you know, managing your vendors. Right? You know, especially with the recent changes to know your customer, kind of obligations that exist out there, there might be more kinds of requirements around you understanding what's going on with your vendors. And so I did wanna take some time just to talk through some of these different kinds of use cases the folks have seen, and I'm sure that folks are seeing their own kinds of bespoke compliance use cases arising, especially when it comes to the realm of government regulations. There are gonna be some regulations that might have incredibly high impacts on your industries that might be relatively unknown to, you know, the broader public. Right? And so although this is a sampling of, you know, the kinds of issues that we can address, know that, generally, as issues arise that require you to have deeper insights into your contracts and business documents, this is a tool that's gonna allow you to address those efficiently and quickly. And so leading off with the kind of vendor management sides of things, you know, a really key part of managing your vendors is looking at things like supplier conduct codes, looking at things like the kinds of terms and conditions that you have in place with your vendors. A lot of times, larger organizations will, as I said, have supplier conduct codes that maybe you've linked in your agreement saying, hey. If you're gonna do business with us, you're going to make sure that you're signing up to these additional kinds of cone codes. And so you're gonna wanna make sure that a high percentage, if not all, of your vendor agreements actually have those codes either linked to in them or as different kinds of amendments in agreements as well. There's also a lot of information when especially as you're a larger company and that you have lots of vendors that you're trying to manage, you might have some that have told and provided evidence, and contractually obligated that they have provided that evidence that they are minority or women owned businesses, right, where in their agreement with you, they have marked that down. But if you're not keeping track of that, when you later go out to public disclosure and need to tell, you know, the government, hey. How many of the vendors are you working with that maybe have that minority woman owned business designation? It's something that might be untracked and you'd be unable to report on. Right? But by actually going into your documents with these vendors and tracking those kinds of certificates, maybe certificates around environmental certifications, etcetera. But being able to track that as structured data, not by hearsay or by folks, you know, voluntarily offering that information, but by actually looking at which one of those have actually contractually obligated themselves to within your agreements is a great way of being able to get that level of visibility and have that level of fidelity in your data. Also things around if you think of, audit and inspection rights, a lot of times, you want to be able to control and audit your vendors, but you don't know if you actually have the rights to audit your vendors. So making sure that your contractual landscapes give you the rights to perform those audits when necessary are important. And, of course, you know, from a government regulation perspective, there could be such a myriad of different regulations folks are thinking about and worried about, you know, from folks on the health care side of the house, thinking more about HIPAA compliance obligations, potentially thinking about physicians that they work with and licensing obligations that they may need to have to perform particular procedures. This whole legal regime set up globally around modern foreign slavery, and so folks working maybe in the fashion or manufacturing industries need to make sure that when they're working with particular vendors, that they have a language in place where they're promising that they're not leveraging any of the kinds of labor that might be prohibited by any of these kinds of statutes. Right? And, of course, there's also so many different types of ESG obligations that have come and gone. And not to mention, you know, the kinds of, you know, data privacy and GDPR regulations, that that have, that have really been a bane many times to folks as they've been trying to, you know, keep compliant in their various kinds of, various kinds of industries. Right? And so you think of a couple years ago, you know, GDPR, has, has particular changes where all of a sudden you need to have additional language in your data privacy agreements. And we have customers all over who are thinking, hey. You know, we have to deal with GDPR. We have to deal with, you know, the California data privacy laws as well. How can we quickly go through and know oh, this particular data privacy is compliant agreement is compliant with this one, but not with this one and not with this third one. Right? And using artificial intelligence. Once again, the structure your unstructured data to be able to answer these questions in more efficient manners allows you to really address problems that even though you're managing thousands or tens of thousands of documents, you're able to quickly get answers from them. And, I mean, audits there's so many kinds of audits that exist. I think one that I do wanna highlight is around cost saving exercises. A lot of times when there's potentially a recession or maybe a corporate, corporate environment where you have to tighten the belt, folks go out and say, hey. How can we save some more money? Do we maybe have, language with any of our, customers that allow us to increase the price maybe by 3% every year, maybe based off CPI or inflation. Potentially, we have rebates in some of our agreements where if we deliver if we purchase x amount of widgets, we'll get certain kinds of volume discounts, and we haven't been collecting on those. Can we go through our agreements and find, you know, those kinds of opportunities? Maybe there's early payment discounts, or or otherwise. Right? But, I I just really think those kinds of audits too are something that we're seeing a lot of. And, Lauren, I know you're gonna show a lot of this in an actual depth versus me just blathering on about it, but there are the other kinds of use cases that you want to highlight. Well, rather than going into more specific use cases, I just wanna emphasize this idea of this flexible model of structuring your data. When a contract comes to you, it's essentially in an unstructured format, meaning somebody needs to go in, read, and interpret what's in it. Well, what we're offering is the ability to very quickly and at scale structure any data point that you need to track, and you're not gonna know right away what all of those are going to be, to Memme's point. There's changes in regulation all the time. There's new requirements for different locations, different governing laws, different, audits that are gonna come up or projects that are gonna come up for your business, then you're not gonna know how to even start that search. So what we're giving you is a very flexible way of very quickly structuring whatever extraction or whatever data point you need at scale across all of your documents. So I will emphasize that more in the demo, but just wanted to make this the point that Memme's making with all of this is that anything can come up at any time, and we're gonna help you get to that data very quickly. Excellent. Yeah. No. That's that's such a good point to emphasize, Lauren. We always say, you're always tracking x amount of things in your documents. You always need x plus one. Right? And with our tool, and we'll show you how, you'll be able to get to that x plus one. So I appreciate that. And, I mean, you must be thinking when I even have that x plus one comment, you know, Lauren's, you know, comment around this flexible system. You're thinking, can I just ask any question and get structured data out of it? And as you're looking at the slide with a list of, you know, many questions scrolling through, the answer really is yes. You know, at the highest level, especially when you want to start looking at custom algorithms to track novel information across your contracts, it does start with the simple question of what would you like to know about your contracts. Right? And this is enabled by capabilities throughout the platform. You know, here we're showing Ask AI, where if you're on an individual document or even maybe a group of documents, maybe kind of in a group of agreements and amendments and such, but you can just start asking questions to the document. Right? Where here, you can see it's asking, hey. Does it refer to the anti kickback statute? Yes. It does get referred to the federal anti kickback statute, etcetera, and gives you the different laws and regulations. And what it says based on in that second box there that's purple and has a number one, that's actually us showing you where in the document things are found. Because, of course, you know, we've built this in a contractual context where, of course, you're gonna need to have fidelity in knowing where exactly this information is. And so we'll enable you to talk to your contracts. And even going further, when you talk to your contracts, you can always have a conversation like you just saw. But as we've been alluding to, you can also just generate novel structured data based out of your questions to these contracts and talking to the AI models. Right? And so, you know, even what we're seeing on this screen here yes. It's a great dashboard, but you'll see it starts with a simple question. How much can the vendor raise the price annually? And then from there, you can save it as an AI model and then have that dashboard that's populating. So as opposed to tasking someone on your team to go review every single one of our vendor agreements and kind of tell me in an Excel sheet, is it 10%, 5%, 11%, etcetera, here you got the AI doing that for you. And like I said, Lauren's actually gonna show this in action, but just really wanna talk to you about, you know, kinda how transformative this is for, you know, different kinds of compliance and audit use cases. Right? And because we've got all types of folks on the call today from all types of industries, we just wanna pull an industry out of a hat, especially because we're talking audits and compliance. We know that, you know, banks deal with a lot of both of those. Right? And so here, thinking of an ISDA agreement, which is, you know, international swap and derivative agreement. You've got clearing and settlement language in there. You've got margin call frequency. You've got information around defaults and terminations. And so here, let's say that, you know, you're worried about termination settlement calculations. Right? You know, maybe due to a particular audit that's coming up and you're wondering, hey. Have how we've been doing our termination settlement calculations, you can just ask the question to the contract and then say, you know what? I actually wanna know this, not just about this document, but across all my documents. You can then save that as a new field across all of your different, documents that's tracked. And then moving forward, you've got dashboards and visibilities to stay compliant and stay ahead of particular risks. Okay? And so, overall, folks, the message that I've been trying to convey is Workday truly does deliver industry leading AI for contracts. And now I think you've heard me talk about it for a good amount of time. I will pass it over to Lauren so she can actually show you some of the technology itself, and I'll be chiming in a little bit as well too. But I'll I'll pass it to you at this point, Lauren. Alright. Thanks so much. Alrighty. Let me go ahead and share my screen here. Alright. There we go. Okay. So kicking off the demo today, I'm just jumping right into a contract that I haven't uploaded so we can get a feel for what you can expect immediately when these documents start to get processed through our system. And then I'll zoom out and we can show you how this data can be used at scale to complete a project or just sift through your data for some kind of ad hoc need that you might have. So starting on this document, I chose a pretty generic services agreement just so we could cover a lot of use cases. And as you can see, it's not the prettiest document. We know that we aren't going to have perfect digitally formatted documents when they hit our system. We want to account for all types of agreements, whether they are brand new PDFs that have gone through a digital system or something that's been sitting in a filing cabinet that has handwriting and coffee stains and, is crumbled up. Because there's no right or wrong way for a contract to live, we account for all quality issues and all file types by running every single document that we upload through an OCR process to digitize all of the text. So at a very minimum, every piece of text and every word and every contract that you upload into Evisort is gonna be searchable and reportable, essentially. So I could run a search for this specific address and have any document where it's mentioned automatically pulled up. Now that's a really great first step, especially if you're used to manual file systems or just a disorganized process in general. But where we take it to the next step is with our foundational AI models that we're gonna apply to each and every contract that you upload. So when Memme and his his, other founders started the company, we leveraged AI for these documents right from the beginning. So for about ten years now, we've been building natural language models to essentially extract information out of contracts. So before AI was even what it is today, it's something that we've been developing and using to ensure that you're automatically going to find key terms and provisions outside of these documents. So here is an assignment provision. Just clicking on that extraction here, and it takes me right to the language inside of that contract. Here's my data breach provision and a force majeure cause. So this list in the middle is showing me a list of different provisions that we have extracted from this contract, leveraging what we call natural language modeling, where essentially we've taught each one of these models on millions of samples of what a force majeure clause is or what a confidentiality clause is. So it has a really broad array of options so it knows how to identify this regardless of how it's written. You'll notice my force majeure clause here doesn't actually say the words force majeure. So that's really important for any outlier contracts that might have different language or some kind of, unique spelling of things. If you were just relying on text based and keyword based searches, you would miss this contract, and you would know that, or you wouldn't know that this is a force majeure, provision that you're gonna need to adhere to in the event of fire, flood, or earthquake. On the far right here, you're seeing some data points that are extracted. So slightly different than the provision models, we're actually pulling out things like key dates, term information, renewal information. So these are more individual data points rather than broad language extracted from that document. Some of these dates or fields are actually calculations. You're gonna need to know when each of these contracts might be expiring. And if you have a automatically renewing contract, that expiration date is gonna change every year or at the completion of every term. So we're gonna actually calculate when your next expiration or when your next renewal is coming up even though it's not explicitly stated because we know what the original date was and we know what these terms are. So that way, if this contract auto renews, you're still gonna have the ability to track the next key date that you're gonna need to manage. Now between these two types of data, so the provision models as well as the data points, there's over 70 different pieces of information that we look for on each and every document that's uploaded. So this includes more than what you're seeing here. This is not a comprehensive list. This is just what this specific contract might contain. From here, we are going to allow you to start tracking any other data point that you might need to find, whether it's right off the bat of uploading these documents or maybe six months into your Evisort environment. At any point, your users or your administrators are gonna have the ability to train new AI models to find unique information out of these documents that might not have come from out of the box. Now why do we do this? To Memme's point, compliance is such a broad term, and it applies to just about everybody. Because we are in the contracting business, every single company or every single organization or every single individual that deals with contracts is gonna have use for a technology like this. So there is no way for us to track every single data point or every single piece of regulation or every single piece of information that you need to track or that all of our users need to track just out of the box. We have to make this scalable. So the way that we do that is actually twofold. One is through what we call a quick AI model, where essentially you can find a piece of information, like here's a standard of care clause. And I can tell the system, I want you to track standard of care language across all of my documents. So I'm gonna go ahead and select that, tag this clause, and give it a name. And now I'm essentially telling the system, learn from what I've highlighted and apply what you learn across every other document that's uploaded. So here, now I can see standard of claw standard of care clause is pulled out in my contract summary. And behind the scenes here, every single document that I've uploaded previously is now being scanned to look for other examples of this language. It doesn't just use the keyword searching like I've said. It actually understands meaning and context and we'll look for other examples of this pretty much globally across the account. Now you do need more than one sample for a robust model, so we would add on to this, with other examples of standard of care language just to make it more robust, but it does start working right away. Now the other type of custom AI starts with our conversational tool that you saw some, slides about from Memme presentation, where I can actually come to a document and ask really any question that I can think of just using a natural prompt that I might that might pop into my head. So what are the data privacy regulations we need to adhere to? Now when I ask this question, it's going to pull from a large language model that we have included behind this tool that only reads and writes contracting language. This is not Chat GPT. This is not a Google Gemini. Those generative AI tools learn from all corners of the Internet. You can use it for just about any type of query you might want, But it's too broad and it's too, open ended to be a workable tool for your legal contracts. You're gonna want something very niche that is only trained to read and write these contracts, and we are always going to cite our sources inside of your document specifically. This is important because especially when you're dealing with an audit or you're dealing with any kind of legal information, you need to be very sure that the information you're giving back from this tool is not hallucinated. ChatGPT style engines have a 27% chance of hallucinating an answer that they're giving you, and that's just not acceptable for this type of project. Also, you're gonna wanna make sure this is a very secure and locked down system. I don't want my contract data being fed back to the chat GBT model. This is a closed system where all we're doing is reading the document in front of you and returning a response with a cited source. So you can be very sure that this information is protected and accurate. It will always give a bit of a broad range as well when I'm asking kind of a vague question like this. So it's being overly inclusive in my answers, but it'll make sure that every answer that's provided here is cited so I can make my own I can, extract the exact piece of information that I need to know. Let me try another one. Is this subject to a most favored nation? And, again, because this is trained on contracts, it should know what I mean when I say MFM. Now I can't ask this tool for a pancake recipe. I can't ask it for the rainfall in Jamaica today. All I can ask it for are contract questions for this document we're looking at. Here we go. It did find a most favored nations clause, but as you can see here, it's actually embedded into a rate change clause. So it knows the meaning of most favored nations even though it's not actually included as part of the language inside of this provision. So this tool is available as just an ad hoc query. I can come and ask any question I need and get a quick response. So for end users or business users or anyone who just doesn't have time to read through a whole contract, find me my answer, and I can move on with my day. Now I mentioned earlier there's this concept of structuring data outside of an unstructured contract. So so far, these answers that I've been given are still unstructured. I asked my question. I got my result. But if I need this information to be structured into a data field, meaning I wanna be able to track at scale which of my contracts have a most favored nation clause or which of my contracts are, beholden to the EU data protection requirements, I can actually create a new AI model out of this query. Very similar to how I built a new AI model out of the standard of care clause, I can do the same with this type of question. This is called custom AI or formerly document x-ray was the term before we were acquired by Workday, where I'm actually going to tell the system, okay. Take this query, apply it to every services agreement, and then give me a response. Maybe it's a yes or no field or maybe it's some drop down field here that I wanted to automatically input over here on the right. So now I can track that at scale and say, show me a list of all my contracts that have a most favored nation clause, or show me a list of all my contracts that have outdated data privacy regulations included. Memme, anything you wanna add to this before I move on to a, a more zoomed outlook? Yeah. I mean, for folks who have been wondering, you know, is it gonna be difficult? Do I need to be a data scientist to train an AI model? These guys are talking about training AI. That seems so hard. How could I do that? Let me show Lauren's about to show you how to, quote, unquote, train an AI model. I think you'll see that it is, quite approachable and very powerful. And so just very excited for you to continue. Great. Alright. So, yeah, let me show you under the hood of one of these custom models quickly, and then we'll go into searching and reporting at scale. So I'm gonna go ahead and pull up a in progress model that's being trained. Let's go with this assignment deviation summary provision. So this is the back end tool that's actually going to allow you to be part of the training of these fields or these outputs so you actually are fully in control of what they look like. A yes or no response is a pretty simple example, but there may be data points that you classify completely differently than what an out of the box field might look like. Do you have a limitation of liability cap, but maybe you have three different kinds inside of your contract and you want the system to go in and pull out the names of each type of liability cap and their respective dollar amounts or what have you, you can get very granular with what you're asking it to pull out. So for this one, we've created a new field called deviation summary. I've given it a prompt that I want it to look through and or that I want it to apply to all of these documents. And then what it's gonna do is in whatever scope of documents I've told it to apply this to, in this case, I've got 430 documents, it's going to give me some samples back and show me where it found the information that it thinks I wanna know. And then I get to tell it, yes. That's correct or no. That's incorrect. If I say that's incorrect, it'll ask me what the correct answer is and why it got it wrong and take that feedback and enhance the model. So this model grows with all of your input over the training period here. And once we have a satisfactory model, I can go ahead and publish it, and now it will be applied to all four hundred and thirtieth scale. So it sounds kind of scary to Memme point when you think about training AI, but, really, you get to be in the driver's seat here. You get to tell it, hey. This isn't actually how I want it to read. I actually want you to change the language. I want you to output just a value. I want you to give me a one, two, three rating score. You are fully in control of that. And, of course, you'll be trained on doing this. We're not gonna just throw into the holes here, but it is actually a fully UI, and user friendly based tool. Yeah. And I think the thing to think about from a compliance and audit perspective is that so much of folks when they think about AI and AI extraction think of, okay. It's gonna go pull out all of my, you know, limitation of liability clauses or all of my, you know, limitation of liability clauses, and then what? I still need to go read it. Right? But what we're talking about here is beyond just the mere retrieval of text. Right? You can summarize things. You can actually even risk score things. So if you say, hey. My sales agreements, I consider them high risk if they have these elements, maybe a termination for convenience, maybe, you know, particular other kinds of elements like rate changes. I consider them medium risk if they have these other elements, and I consider them low risk if it's these last set of elements. You can then set loose the AI on all your documents, And it's not just going back and pulling back, oh, this is your assignment. This is your termination clause tax. It's actually telling you, hey. Based off how you think about risk, these are your high risk document sales agreements. These are your medium risk, and these are your low risk ones. Right? And so, I just you know, in thinking about this kind of, configurability, the configurability is important because you have novel issues, but the configurability is also important because even when you have similar issues to other organizations, you might have different risk tolerances. Right? You might have different ways that you address that particular kind of risk, and the AI is gonna be flexible enough to meet you however you want to kind of, manifest your approach to risk. Just wanted to find one of the examples of the risk score provision. So here's one we made about data privacy risk. So we're asking the tool to actually classify these documents as a high, medium, or low, and then we are the ones telling it what those mean. So for certain areas of the business, this might have completely different meaning than another. So you can get really specific here and say, okay. It's a low risk if we have ISO twenty seven thousand zero one protection. It's a medium risk if it complies with GDPR, but it's a high risk if it's vague and we don't actually know. So then it interprets what you've written here and will make will make educated I don't wanna say guesses, but educated extractions from all of these documents to ensure that all of your documents have the appropriate risk score applied. So just one way that this can be used, but, really, there the the limits are endless. Indeed. But, yeah, none cutting in. Please continue. Yeah. Absolutely. So now that we've looked at both foundational AI, what you get day one, and that well, we haven't mentioned this, but when your documents are uploaded, those extractions are made almost immediately. It's a very, very quick process to turn those documents into structured data using our out of the box models. We've talked about that. We've talked about custom AI. Now I wanna get into how do we use that information at scale. So it's great if I'm just looking at a singular document, but what's more important is I need to dig through everything, and I need to find things that are risky, or I need to find things that fall into outdated regulations or whatever it may be. And that's where this reporting engine is gonna come into play. So I'm gonna start with just a blank search and pull up everything we have in this repository, which is 8,700 documents that we've got here. Any of the data points that are extracted out of these contracts, out of the box, custom, even manual data fields. So if you just wanted to tag your documents with manual data, that's fine too. All of it becomes reportable at scale. So I'm gonna go ahead and add some other filters here from some of these fields. We're gonna use the let's see. Let's do do the data privacy, assignment risk. I'm essentially just telling the system, I want these to be something I can search on. Here here's our risk score. There we go. So now I can say, alright. Show me all my contracts that have a data privacy risk of high. I'm gonna filter on that data point, and it's now gonna narrow me down to a 121 documents. Great. This is something that we're gonna need to pay attention to. Maybe we want to audit all of these and renegotiate them upon renewal. So I'm also gonna add another filter saying, show me anything that is renewing of this list in the next six months. I've got four. Okay. We need to now take a look at these and see what we're gonna negotiate differently upon renewal. So I'm gonna go ahead and select all of these, I'm actually gonna assign them to somebody. So we're gonna give these to an owner. So my contractor viewer is going to be Hailey Rockins here, and he is now assigned as the reviewer of these four contracts. Ryan might have his own alert set up or his own reports set up that notify him whenever he's assigned a new contract. We've just added four new tasks to his desk. But, really, any of these data points can be turned into some kind of automated alert. Maybe I wanna know every time a new high risk contract is entered into the system. I can have the system automatically send me an email saying, hey. Here's a new contract with a high risk data privacy score. Pay attention to it. Or using a, more broad example, any kind of date that we're tracking is also gonna be really important to make sure is in front of you on a regular cadence. So for things like services agreements or client agreements, I wanna know when things are renewing. So show me everything that's renewing in the next three months here. And then of these, maybe I wanna see everything that is renewing automatically. So talking about paying attention to make sure things aren't renewing that you're not using anymore, any outdated services, or any kind of noncompliant system that we might be using. I wanna make sure those things don't auto renew. So I'm gonna have an email automatically generated to me every Friday or at the end of every quarter with the rolling list of all of my renewals just automatically sent to my inbox. I don't even have to come into the system to find them. This is also great for all types of Boolean searching. So we mentioned at the beginning of the demo that every contract is scanned through OCR. So that's gonna be really important if I need to find things like a name or I don't have, a model trained right now looking for GDPR language. Show me every contract that mentions GDPR. And I'm just doing a Boolean smart match search, so it's looking for this exact language in my contracts. And now I can actually start auditing this list directly inside of this report. I don't even have to drill into the contracts to find both where the that those keywords might have been mentioned, but I can also see what these contracts contain directly from here. Another option for reporting besides these ad hoc generations of reports or an alert that I'm automating this this, data to be sent to me, we also have dashboards. So this is gonna be, one of the slides that Memme mentioned where we're actually turning this data into a visualization tool. Now, again, any of the data points that we extract, whether it's an out of the box field, whether it's something custom, can all be turned into a dashboard. And it's really easy to do. So I'm just gonna go ahead and select a few here from this list, click apply, and it'll automatically add those to what I'm viewing here. I can also add filters. So if I only wanna look at a dashboard for specific contracts, I can narrow down the scope of what I'm looking at, or I can even just use these to drill into. So if I wanna look at credit agreements, I can just select that bar on that chart. All the other contract data fields are going to update, and I'm gonna have a list of those contracts down here to work off of. These can be configured however I want. I can remove any data point I don't care about. And most importantly, I can create as many dash boards and share them amongst my team members that I might need. So here's an audit report management dashboard that Riley created that's just looking at audit report cadence or audit report types. I have another in here for data privacy regulatory compliance. Here is that data privacy risk score that we've already looked at a few times or regulatory requirements that we need to adhere to. Breach notice is going to be one of our out of the box fields, but it's really important to understand in the event of some kind of a in some kind of, breach or something else that you need to notify all of these contract holders about, I wanna know what my obligations are and how soon I need to get that information in front of those customers or those clients. Again And I would only I would only add here is that, I mean, you look at that data privacy risk score, but these are dynamic dashboards. So you see that big bright yellow high risk? As a lawyer, you just wanna click that. And now you are only focused on those high risk documents. You've got the list of them at the bottom, and you can click in. If there's another red flag dashboard, you could click that there. I just did wanna have you click one so folks can see. These aren't just static dashboards. This is living, breathing document data that they're engaging with. Absolutely. Going back to the idea of the flexible structuring of data, which I mentioned, a couple times now. Really, the whole takeaway that I want you to have from this is that the data is there in your contracts. You bring them to us in an unstructured format, whether that's a crumpled document, whether it's a clean PDF, whether it's a picture of a handwritten napkin that somebody signed over coffee, whatever it may be that qualifies as a contract, you can upload that into our system and we are going to structure it for you and give you whatever output is going to be most useful for your day to day. How long have you spent reading and digging through contracts trying to find what it is you are obligated to adhere to? Or you have an audit coming up and you're panicking because you don't even know where all of your documents live. Now we're just offering a single source of truth and a very flexible model to find whatever data point you need at scale and very quickly. So I hope this was helpful. We have about ten minutes left, so I'm happy to just open it up for general q and a. The chat is open, and we have quite a few moderators behind the scenes that are here to answer questions. So, if anybody wants to, put their questions in there, we'll make sure we address them. Memme, any other closing thoughts from you before I pass it back? No. Feel free to pass it back. That's totally fine. And I'll put the we just got some closing things we wanna kinda show folks on on the screen beforehand. Folks, if you don't have questions, maybe just let us know what's the compliance and audit issue that you're, dealing with, you know, right now. You know? Are there if there are any, is there any of those? Yeah. I will actually read some of the q and a to you if you want to answer them. Okay. Yeah. That'd be good. I'm I'm actually not seeing them, so that'd be great. Okay. Great. So we have a question. Can the query results be exported back out to where the contracts live, I e Salesforce? Yeah. Most definitely. And so, one of our key capabilities is integrations. And when it comes to our Salesforce integration, we actually allow not only the ingesting of documents where we're pulling in your key documents from Salesforce, but then once we analyze those and create the structured data that you saw, we can push that data back into Salesforce where there would actually be in a little box with, you know, metadata that you care about from that perspective. Excellent. If our contracts all reside in an email inbox, can you extract them from there? That's a really good question. And so, one, we actually even have an email address you'd be able to send it to to get them all in there. But, frankly, as part of the onboarding, just know that we're your partners in sniffing out all your contracts and bringing them into the tool. Because as you saw, a lot of the times, you might have scanned PDFs. You might have images of documents. But when we say we can analyze 200,000 documents in under twenty four hours, that's every kind of format of document, be it an image, a PDF, or a Word doc. And so we're gonna be your partners in finding all those documents, be they in email and box and share point some people, have them just in DocuSign because that's kinda where that's their repository of the executed agreements. But as a part of that onboarding, we're bringing those documents in and giving you that data. Right. And the question of the hour, how does this interface with strategic sourcing that we are already using? Okay. Lauren, did you wanna dive a little bit into that one? Sure. So Evisort was acquired by Workday about eight months ago, and we've been slowly tying the, you know, our original company into the Workday ecosystem. As of today, it is still its stand alone product that we showed you on the demo. But our first point of integration is actually coming into the strategic sourcing tool. That's the first place that we're actually merging the two platforms with an integration. So, the workflow engine that we actually didn't show in the demo today is going to be a part of the strategic sourcing module going forward. So it's actually enhancing what you're able to do today with strategic sourcing contract by adding a more robust workflow and AI layer to the negotiation of your procurement contracts. So there's more to come on that. If you are interested, I would say reach out to your Workday rep, and we can show you a demo of what that's going to look like. Awesome. Excellent. And I fixed my thing. I actually can see the questions now. Continue to feed me on if if you got some more. Yeah. Absolutely. It does look like some of them have already been answered by our moderator moderators backstage. However, I will, go ahead and just read some anyway. So, with the change of leadership in legal opinion or risk profile, what would you say is the best way to adjust a trained model that may have been driven by previous leadership? Create and train a new model or retrain an existing one for a given provision? Nice. Yeah. And it looks like, we've responded that, you know, the simplest and cleanest way would be, you know, the former, especially the fact that it's a pretty low lift to create a new model, and then you can kinda train them specifically around new parameters in a kinda clean state. And so, yeah, it's it's really a living and breathing kinda tool. And you saw it's made to be a it's made to be engaged with. Right? You don't have to learn how to do aggression models to train the model. You just need to have an understanding of what you're looking for. And even better, and this is something we've been finding as, you know, really rolling this out to customers big and small, if you can understand, hey. This is how I want it to give me answers. Maybe it'll just go back and give me the exact clause text for everything, but I wanna understand it as high risk, low risk, or medium risk. Or maybe I wanna understand it as either it's this type or that type, x or y, so you have a really clean kind of pie chart at the end of the day, I think, is really important. As we do wrap up here, I do wanna give some folks some takeaways so they can kinda continue kinda learning and thinking around us. A little bit of, I guess, a self plug in that I do cohost, you know, the legal AI podcast work that has meaning of the minds. We talked to so many thought leaders at the intersection of law and AI. And if you want to hear some more conversation like this in those podcast formats, please do scan that QR code. And, also, if you have even broader questions, you can actually get a personalized Workday demo by, of course, reaching out to your Workday account rep. But we actually have a little bit of a link down there for folks to potentially use as well if that's what they're looking to do. And, you know, now that you've seen this, we are, like we said in the beginning of this video, gonna have a recording of today's session available on demand. And we really do think, though, if you have questions about Workday CLM, just email, call your account executive to get your questions answered and learn more. This technology is really incredible, and it kind of is too good to be true, but it is true as a lot of the exciting innovations happening around AI these days. And so, definitely, if you're curious about it, learn more, look at it for yourself. And then most importantly, you know, learn keep abreast of these kinds of innovations by registering for upcoming Workday live sessions by scanning this QR code. Folks, it's been an absolute pleasure. Lauren, thank you so much for that incredible walk through of so many different features and capabilities of the platform. It's been a pleasure. Thank you so much, Renee. And thank you everybody who joined, and we hope to hear from you soon. Awesome. Excellent. With that, I think we'll give everyone, I guess, about one minute back. Thank you so much for your time, and enjoy the week.