Video: Live Lab: Master the latest Agiloft release with product expert tips & tricks | Duration: 1492s | Summary: Live Lab: Master the latest Agiloft release with product expert tips & tricks | Chapters: Welcome and Introduction (27.64s), Screens Overview Demo (89.225s), Screens and Obligations (193.745s), Exploring Screens Community (777.15s), Custom Extraction Process (856.8s), Configuring Obligation Extraction (939.94s), Obligation Extraction Demonstration (1067.755s), Q&A and Conclusion (1197.59s)
Transcript for "Live Lab: Master the latest Agiloft release with product expert tips & tricks": We should give it another minute or so, and then we'll get started. Oh, I think I can see people signing in now, which is great. Okay. So hi, everyone. Welcome to this community live lab, where today we're gonna walk through how to use screens, for both pre and post signature workflows. So for those of you who did join our release webinar last week, you would have seen some short demos. And today we're gonna go into a little bit more detail and show how these are set up and configured in a knowledge base. So my name's James. I lead one of our product development teams here at Agiloft. And today, I'm joined by my colleague Marissa, who's going to kick things off for us today. Marissa, over to you. Great. Thanks, James. Yes. I am a product manager for Screens. I am focused more on the presignature side, and that's what I'm going to walk through my demo. So we're just gonna start with a quick overview of what is Screens. So if we can jump to that slide. So, as many of you know, Screens is a AI based contract review and analysis tool, and it spans both pre and pro post signature functionality. So, some of our, presignature features are focused on generating playbooks to accelerate contract review. So we have a big suite of community playbooks, which are already built for you if you wanted to tap into those, or you can build your own as a set of pass fail requirements for your contracts. And by applying those in our Word app, or even running it straight from the Agiloft KB, you're able to quickly assess your contracts. And specifically in the Word app, you're able to redline based off your preferred language and your default comments. There's also a suite of post signature functionality that's available through screens, and that's what James will walk through a little bit later. And that's really focused more on contract analysis, specifically the ability to detect and track key information, including obligations. So if you're wondering if you are eligible for screens, if you're on one of our legacy editions and you have our AI platform add on, then you do have access to screens. And if you are on our new AI in the inside CLM edition, you have access to screens. So if you're unsure or if you want to make that transition to the new edition, I encourage you to reach out to your account manager or your c or your CSM. We'll be happy to get you on the AI activation pathway. So now we're going to actually step into screens, and I'm going to give you a bit of a demo. So starting with what I mentioned of how do you even know if you have screens. So we're gonna start by, if you're an admin, going into this setup section and going under integrations, and you will see right there that screens is deployed. You can also check by just searching for the screen results and screen run tables. If those are present, you're all set for screens. So as an admin to hop in, you could just click this build screens, and it will have you log in using your KB credentials. You also could log in directly on that app.screens.ai website, which is accessible to anyone just using their KB credentials as long as you're a power user. So when you get in here, you wanna start by building a screen, and you'll see the four different options up at the top, whether you're generating from a contract, using a community screen, or building from scratch, or if you happen to already have a screen from another environment, you can upload it. I'd recommend if you're getting started just working with a community screen, and that's what I did here. I'm gonna pop into this one. I actually just duplicated a community screen. And this is from, the perspective of I'm a contract manager. I'm helping our products and engineering team assess pretty low value SaaS contracts for tools that they're using to build. So any playbook is going to be made up of standards, which are pass fail requirements, and questions. And questions are just more specific information or details you want, like what's the governing law. We're gonna focus on standards right now because that's really the meat of it. And what these are are just individual requirements for every topic. So you're seeing some here like data liability cap, third party indemnity. I'm gonna pick a simple one, so audit few restrictions, and we're gonna go in to edit the configuration so you can see what this entails. So this whole concept is the idea that I don't want this vendor to go assess usage overages and just start charging me for fees without our consent. So in the standard configuration section, I first need to start with the standard text, which is pretty succinct one to two sentences saying what your requirement is. Then if you wanna add standard details, this is just a little bit more information. So you could provide some examples, a little bit more context. All this is just a prompt that's going to be sent to assess whether the requirement passes or fails. There's also this non address section, which is focused more on if this does if this requirement isn't mentioned in the contract. Sometimes you want it to always pass, like for us. Sometimes you'd want it to always fail, or you could just have the AI infer it, and that's what that provides. This redlining configuration section is new in release 32, so it's November. Preferred language is what you ideally would like to see when you're redlining. So this is your preferred contract language. If you have a clause library already built in your KB and you have preferred language already there, you can go ahead and search it. I'm quickly using the filtering feature, and it can bring me specifically to a clause that I already have on audit fee restrictions. So if you have that and wanna select it, you just click submit and it paste it right in there. So it's designed to just speed up that process instead of having to flip and search within your KB. When it comes to redlining approach, that's determining how heavily you wanna rely on that preferred language, either surgically where it's more lightly using it or heavy where it's as close to verbatim. And then similarly with the default comment, this is if the standard fails, what comment do I wanna appear in the word add in? And that similarly can be AI adapted or just a copy paste verbatim comment. So that's what I've added in this section. Down at the bottom with the risk configuration, that's assessing your perceived risk if this standard fails. So for something like this, I'm just gonna switch that to medium and say it's moderate impact. And then user guidance, it doesn't affect the AI. It's just for users to understand how to use this standard. So once you save that, you would need to build this for each of your standards. But, collectively, that will generate a playbook that then you can then put into use. So if you flip back into your KB, I have right here a SaaS agreement, a terms of use document. And I'm gonna open up that contract record. I'm gonna go down to the attachments. And it's really important that I check out to edit from the KB because that's going to allow the data from screens to get pushed back into the knowledge base. So now we're flipping into the word add in, and I'm logging into screens. And what we're gonna do is run that screen that we just built. So I'm gonna find this low value SAS contract one run one and run it. So you will get the results for every standard, and you'll see that they either have a pass or a fail result. I personally like to use the filters at the top because I like to jump right to the fails since that's where the focus will be for redlining. I also like the search, so I'm gonna jump to the audit fee restriction standard that we just configured. And if I click on it, it brings me to the relevant language, and the details will explain why there's a fail here. So I have the AI reasoning, which is saying, hey. We didn't want to just be charged overage fees without consent, and that's exactly what this document says. It also cites the preferred language that I indicated, and it has used that to shape the red lines. Because I had selected heavy red lines, it's used it pretty verbatim. I also have my comment, which I similarly had asked to be verbatim, and so that's there, and then the user guidance as well. So if this is satisfactory, you can click apply red line. You'll notice the red line is applied. The default comment is also there. But oftentimes, you don't wanna do it one by one, so we have this resolve all fails with one click, which if you just go ahead and press that button, it'll just take a few moments, but it will apply the red lines for every failed standard. So what you'll see here is 10 red lines were successfully applied. The one other thing I wanna show you is that if you jump back to the KB and we go to the same document and we move to the screen results tab in there, you'll notice that all of these results have written back to the KB. So for that screen we just ran, we have every element name. We also have the risk level that's associated with each of these, and we have the results of the pass fail. So in addition to being able to see this in the word add in, you can see it over in screens. Now, obviously, I focused on presignature use cases. I'm gonna hand it over to James, and he's gonna show you how you can do even more post signature, within the KB. Great. Thanks, Marissa. Right. So, okay. Just a quick reminder for everyone that's, with us today. If you have any questions, just please pop that in the chat. I mean, there was a q and a function, in the chat there, and we'll come back to those at the end. So as we said, I'm now gonna talk about, what happens after a contract signed. So, how you might use screens, for that use case. So I'm gonna talk about obligation, our new obligation extraction feature. And this is powered by the same AI service that we're using for screens. And, really, this feature is aimed at any of our users out there who might be responsible for reviewing signed contracts. Maybe you're identifying the obligations within those contracts and perhaps even assigning those out to relevant teams. Now I imagine if you are doing this exercise today, then this likely involves you reading contracts line by line, copying and pasting sections into either an Excel spreadsheet or even a Word document. And this takes time. Right? Like, especially if you're having to review maybe more than one contract in a given week. So our new obligation extraction feature not only speeds this process up, but we're also able to maintain a really high level of precision to help deliver concise, accurate summaries of your most important obligations. And so that means you don't need to go digging through pages of text to find the answers that you need when reviewing your supplier contracts, for example. So how does this work? Our team of legal knowledge, engineers have built a set of obligation screens that are now readily available in the screens community. And they cover eight individual topics, around 70 subtopics, and also three agreement types. Now for those that don't know what the screens community is, don't worry. I'm gonna show you that in just a second. So once you've created your chosen screen, perhaps you've even created your own screen, that's when you can then use the new screens run action to execute on that contract. And this will automate the extraction of obligations. And then we use a set of rules and actions to store that data in a new set of tables that we deploy into your knowledge base. Alright. Well, sounds pretty cool. Right? K. Let's see it in action. Okay. So we're starting over in, our screens community. And the screens community essentially provides a library of prebuilt screens from a community of users and experts, which includes our very own legal knowledge engineers. And so this provides users with access to deep industry know how and expertise across a range of your contracting needs. So things like m and a and finance, maybe it's privacy or real estate. And we also have a section in here for obligation management. So we're gonna go ahead and take a look at an example here of this SAS agreement screen, which when you click into, we can provide a bit of detail about what that screen's all about, who the author is, and then we give you the option to go ahead and grab that screen and download that directly into your integrated screens workspace. Now once I'm in the screen, I can see that it's made up of a number of questions. And each one of those questions provides a specific prompt. So as a user, you can look at the detail of each question. You can see the prompts. You can see what the scope is. We provide complete transparency about what that prompt's doing. And what makes these screens unique is that we're using a consistent format to structure the output of each extraction. And as long as this structure is maintained, this format maintained, these extractions will automatically import, into the obligations table in your KB. Okay. So now I just wanna show how easy it is for a user to go ahead and maybe add your own, custom extraction. So each screen contains a template extraction at the bottom, which I can either edit directly or make a copy. And let's say I want to know about, any post incident obligations following some kind of breach. And I wanna know what actions my supplier, needs to take. So I'm gonna go ahead and just add a new question into the question text. I've prepared one. I'm just gonna copy and paste that into this section. And, note that it's also important to add a valid obligation type here at the bottom. So in this case, I'm gonna use categorize that as performance, and that helps us to categorize the obligations when they come back into the KB. Of course, you could add, any additional detail to any one of these data points, and, we provide a bit of guidance here at the bottom on how to do that. But, to keep it simple, we're just gonna add a question and the type. So then we click save, and now this, screen is ready to go and ready to be used in KB. So I'm gonna switch over to the KB now, and you can see, we're viewing an obligation extraction dashboard. And this gives me some useful insights into data that's already been extracted using screens. So if we jump over to our contracts table, and I select a contract record for my supplier SaaS agreement, we're going to go ahead and extract the obligations from this attached document. So you'll see, I have a tab here specifically set up for obligations and I'm able to now choose a file. And when I press this extract obligations button, this is what's calling the new screens run action in the KB, and then I'm asked to choose a screen. So I'm gonna go ahead and select the SAS screen, that we just worked on. Now that's gonna take a minute or so to run. So let's just go take take a look at how this button's actually configured. So I'm gonna go over to the attachments table. And so it's important to note that the screen's action is only available on the attachment record. And once that screen integration is deployed to the KB, then this screen action will be available to choose from from the list of actions that we see here. So I'm gonna go ahead, create a new action. We'll just give this action a name. And then there's a couple of other fields that I just need to go ahead and fill out. You could give it a description if you need to. You need to choose where the source file is coming from, and then you need to select, how you wanna choose that screen. And you've got three options here. You could either pick from a list. You could configure it from a text field if you wanted to preconfigure something, or you could make a selection to choose it at run time. And that's, the option that we just saw earlier where the user has to to make that choice after they hit the button. And that's it. So that action is then created, and that can now be assigned to any action button on the layout. Okay. So we're gonna switch back now to, the contract record. And if I open that record where we ran it earlier, go back to the obligations tab, we'll now see extracted summaries for each obligation. And this will help us to answer some essential questions about this contract. So things like what are my payment terms, maybe what's the agreed availability uptime each month. And if we scroll down right down to the end, we'll see that custom, question which we included in the screen. So are there any post incident obligations that I need to know about? Now, of course, this is just one simple example, but you could add any number of these custom extractions, add some more details to that prompt using that guidance contained within the screens template that I showed you earlier. So let's take a look at one of these obligation extractions in a bit more detail. And what you'll see is that Agile not only extracts and summarizes the relevant text, but we're also categorizing each obligation. We identify any relevant triggers and also responsible parties. And the coolest thing we're doing here is to create that structured metadata. So that makes working with data much easier because you can now report on that data. You could even send it to an integrated system. And, of course, no obligation management solution is complete without the use of tasks. So we're also able to link that extraction to our existing task management capability so users can start to, you know, keep track of ownership as well as deadlines. So with obligation extraction now, we can deliver precise, actionable summaries in just a few clicks so users don't need to go wasting time searching through files. And at the same time, we're able to maintain accountability once those obligations have been assigned. Okay. That, brings us to the end of our demos for today. I think we should have some time for any questions. Marissa, have you have we James, got any if you came for you while you're presenting, the first on extracted obligations, which table slash what obligation type is default for the output? questions? It's default for the, which table what obligations? So the table, we we're creating a new table, essentially. We're gonna have an obligations table, which we, is part of the new CLM template, which went out with our release last week. And we'll also have a deploy method available to for existing customers, and that will deploy a new set of tables into the KB. And that should be ready to go probably in the next release. Great. And this one came while you're presenting but also spans both pre and post signature. Which agreement types are available and what new types are on the road map? So maybe if you wanna speak on the obligation side. So agreement types or obligations. I think we've got one for, like, a general, general commercial agreement. We've got one for a Saks agreement, which is what I showed in the demo. We've also covering one for a statement of work. And then we can build a road map based on sort of customer feedback where once we, you know, if we get requests for types of agreements, we can take that back to our legal knowledge engineering team and and quickly spin up new more types. Great. There are a few that I can answer here. So does Ask AI come for free with a standard Agileft license? So if I'm interpreting that you're using the word standard to mean, we have a admin standard and requester licenses. That implies that you are on the new AI on the inside package. And if that's the case, yes, Assai is included in that. And the same thing is with screens as well. If you're on AI on the inside, then you are entitled to this functionality. And then there's another similar question about the overlap, specifically about the chatbot, which is correct. There is a chat both in the word add in for screens itself, and then there's Ask AI. Does the usage depend on user in context? I would say that it is, focused largely on what you're trying to achieve. So for the ask, for the chatbot that's in screens, you know, it's just looking at that document, and it's just going to persist for that particular session that you have there. So to me, that's more of a quick query that you might have about the document while you're actively in the negotiation stage. Ask AI is more robust and expansive in that, yes, we can you can do a single document chat, but you also can do more comprehensive chatting, so for instance, within a contract record itself. A few other questions. Can you run multiple screens from the KB, for example, when you're running obligations extractions? Yes. So the screens action will allow you to select, multiple screens. I think you could select up to about 20 screens in a single action. Wonderful. And then a few more for me. Is there a limit on the number of standards that you can add to a screen? So there's not a technical limit, so you can add a large number. But for optimal processing, we'd recommend maybe, like, 20 to 30 maximum and and perhaps break up your screens by topic or by risk. The biggest challenge that you'd run into is when you use that one click, red line solution that I demoed because there's a risk of just a lot of overlapping standards and overlapping edits if you have, like, hundreds of standards in place. Those are all the questions we received. We can give it another moment to see if any come into the chat. But, otherwise, this is all recorded, and, the marketing team will send this out to you. And we also highly encourage you to keep the conversation going on community, especially if you have feedback about anything related to screens or the community screens themselves. Great. Alright. Well, thanks everyone for your time and for joining us today. And, yeah, we'll look out for the email with the link to the recording, and we'll hopefully see you again soon. Thank. you, everyone. Thank you, James. Thank you. Thanks, Marissa. Bye.