Name: Agiloft + Screens: An integrated experience
Uploaded: 2025-04-30T17:01:04.813Z
Duration: 39 min 44 s
Description: Agiloft + Screens: An integrated experience

Transcript for "Agiloft + Screens: An integrated experience": Hey, everyone. We're gonna give everybody just a couple minutes here to, get logged in. Alright. We still have some folks strolling in, so I'll give it just another minute or so. Alright. Good morning or good afternoon depending on where you're joining us from today. Thank you so much for joining. My name is Anne Pollitt. I am a product marketing manager here at Agiloft, and we are extremely excited, to have you join us here today for Agiloft in screens, an integrated experience. Just a few things before we start. There will be a recording available after this webinar is complete, that you can access at any time on demand. Also, the chat and the Q and A are open, and so there will be a time at the end of the presentation for questions. So at any time, throughout, if you, have any questions, please feel free to drop them in there, and we'll go ahead and get started. So as I mentioned, my name is Anne Pollitt. I am joined today, by two of our great product managers, Jack Wicks and Nate Ruiz. To start here, Jack, to know him and still love him. He's been with Agiloft a long time. I'm sure all of you, or a lot of you at least have had the opportunity to work with him. And then, Nate Ruiz who, after our acquisition of screens, joined our product organization in January, and we are extremely excited to have him. Gentlemen, do you wanna go ahead and introduce yourselves? Yep. Hi, everyone. I'm sure most of you have seen me on these webinars before. I'm joining today to help out with the q and a at the end. I've been working very closely with Nate and the rest of the team that came over from the Screens acquisition over the last few months to bring this integration to life. I'm really excited to share it with you all today, and I'm excited for you all to meet Nate, and I'll pass it over to him now. Hey, everyone. Thanks for being here. Again, I'm Nate Ruiz. I came over with the screens team as part of the acquisition. And like Jack Wicks mentioned, we've been hustling to, integrate screens with Agiloft. And on the call today, I'm super excited to be able to show you that the first step in that integration. Thanks, guys. We're happy to have you. Alright. So before we jump into our demo, let me give you a little bit of background about some stuff that's changed. So there's been a lot of exciting news out of Agiloft over the last year. We have had a lot of amazing new features, especially on the AI front, had that have been released. We have, we announced the, investment from KKR. That was super exciting. And then, of course, as we mentioned, the acquisition of screens early this year. And so our teams have been working diligently to get that powerful technology, completely integrated into the Agiloft experience. So, for those who have seen in the e in the emails, maybe even, in the news and and via press, that has happened. So screens is officially integrated into the core Agiloft platform. Anyone who is using, the AI platform today already has access to screens. So, we're very excited for them to get to see what that means today. And then, ultimately, that came, about with another large change that hopefully you've had the opportunity to see, and that change is AI on the inside. And, essentially, what has happened is we've we've really simplified both our packaging and our licensing structure, in order to make the AI features more accessible. We know that that's the way of the world. That's the way the world is going, and our teams have worked super diligently to create thoughtful and practical ways to use AI in contracting. And we wanted that to become be accessible to any and every user. So but that what does AI on the inside mean? It means an AI enhancement at every meaningful meaningful step of the contracting process. We're gonna take a look at what the process looks like, at least according to, Agiloft and and kind of what that means to be AI enhanced. But we wanna continue delivering on our differentiated value to all customers. Right? We know that these tools, are already being used by certain customers today and making a great impact, and, ultimately, we wanna share that with the rest, of our organization. And then now ultimately included in our core package. So, for those of you that have questions, about what that new package looks like and what those new licenses look like for you, please do not hesitate to contact your account manager or your customer success manager directly. They will have a lot more details for you. But also keep an eye out in your email. We are going to be having another webinar that's focused on what AI on the inside really looks like, an end to end, enhanced contracting workflow with AI, and that'll happen, towards the end of our end of this next month or early June. So please keep an eye out for that. Alright. So some of these you'll be familiar with. Some of the names have looked a little bit different, but let's look a little bit closer at AI on the inside. So contract analysis, you might have seen this as third party analysis, or a few other names. But what this really is is being able to use AI to find important metadata from an unstructured contract, and be able to ingest that and then use it in your contracting workflow. Reviewing negotiation, I'm not gonna get too far in-depth in because Nate is going to show you all that magic in just a few minutes. And then Ask AI. So you might have seen that as Convo AI, document or repository q and a, if any of those names sound familiar to you. That is also now included in the core package. So either legal or team members beyond legal can ask questions of a contract and and really become more independent, in working throughout the contracting process. So that's super exciting. What you see on the right and highlighted in orange, prompt lab and then also batch analysis or previously referred to as batch import, those are still gonna be, add on pieces, but we still know are extremely powerful in the overall, AI workflow. So we just wanted to highlight those. So a quick look at what that contracting workflow looks like. So, again, I I think it's really important to remember what makes Agiloft unique and and why some of these even some of the questions we'll probably get today are gonna be different for different members. And that's because, you have taken the time to make Agiloft work for you. Right? We at Agiloft, we know CLM, and we know it well. And so we've worked hard to, really perfect ways to use AI within CLM, but you know your organization best. And so we want you to be able to pick and choose where that happens, where AI can be used. So, for example, we have current customers that use it as a preliminary risk analysis so they know where exactly to write route contracts, who's in who's, supposed to take the next look in the past. We have customers who, want to use it as regulations are changing and updated, in the industry. They want to see which contracts immediately need their attention. Again, we definitely have customers using it for review and negotiation, on the screen side, so can't wait to take a look at that. Maybe a summary towards the end of a contract, to set to a financial team. The options are really endless, and so they're there for you to use just like other features within Agiloft where you think is best and where might makes most sense for your organization. So we're very excited about that. Again, a lot more to come on AI in the inside, what that means for you, your organization, what the new packaging and pricing and licensing means. Definitely a lot more to come on that. But please, in the meantime, contact your account manager or customer success manager directly because we are extremely, extremely excited about these changes. Alright. So without further ado, I would love to get to the star of the show. Nate, take it away. Perfect. Thanks, Anne Pollitt. Alright. So today, I'm gonna give a pretty high level overview of, Screens itself as well as our, you know, first integration with Agiloft. Screens is, Screens is a contract review and negotiation tool. And after the recent, acquisition by Agiloft, we've now started to move more and more of our functionality, into Agiloft so you can access Screens and all the power it brings. So screens is, accessed through a web app or a Microsoft Word add in. We're gonna start in the web app because that's where, screens are built and operationalized. So this is the screens web app. You can see up here at the top that there are a few different ways to get started with the screen. We try and make it as easy as possible on the screen side for you to get up and running with a functional screen that you can then continue refining as you're renewing contract. So at the top here, you can see that you can actually generate a first pass screen from your standard position contract. We have a community of of screens that are published by contract experts, And you can this community is actually publicly accessible, and you can come in here and grab any of these screens and build off of them. So for the purpose of this demo, I'll go ahead and download this SaaS savvy lower value purchases screen. Oh, I've gotta sign in. Okay. So that screen has been added to my account, and you can kind of start to see the structure of a screen within the web app. Again, the web app is gonna be used primarily for building and testing your screens. Most of these screens will actually be operationalized within the word add in, which we'll jump into in just a moment. So this screen that I just downloaded is a community screen. It contains 24 standards and five questions. So standards are pass fail requirements that you want to be met when you're negotiating a contract or when you're reviewing a contract. We can look at the first standard here and it it says the contract should require that the vendor commit to complying with all applicable law laws. That's a pretty plain plain English requirement. And when we run the screen on a contract, we'll be sending this standard along with all the standards in questions to the LLM and we'll be getting back whether the contract passes or fails the standard. We can jump into the configuration to see kinda just how easy it is to build out one of these standards. Again, it's it's plain text the, creator just typed the contract should require, and then we give you a few extra, configuration options. You can designate a risk level, and we'll use that risk level to calculate a weighted score for the screen when it's run on a document. We allow you to prompt the LLM directly, additionally, here in AI guidance. So we recommend the high level prompt be super clear so you know exactly what you're testing for. But if there are addition if there's additional context that you wanna give the LLM or clarification, you can include that in AI guidance. User guidance is used primarily for communication with the end user of this screen. We've seen customers use it a handful of ways. It could be to advise on, you know, an approver that needs to come in and approve a given standard. It could be on the standard itself and and why it's important to a business. But this user guidance, config is is one that you can use however you want, to communicate with the end user. If we jump in and edit, we'll also see an option for preferred language. We give we give users the option to paste a preferred clause right here in line. And when a user is running the screen on a document, they'll see that preferred language right in line. So that's standards in a nutshell. I'll briefly touch on questions. Questions are prompts that help you extract specific information or or data from documents. So this community screen creator has created five questions, and these are usually things that you might wanna know about a contract, but you don't have, a strong opinion about or you you don't you don't really care if it passes or fails. So in this first one, you know, governing law venue, what is the governing law and venue for dispute resolution? And then within the user guidance, this this, community screen creator has added some some education for the end user. So that is a screen, and now we'll we'll show you how we operationalize these screens, from an Agiloft contract review workflow. So I'll hop into Agiloft and come to my contracts table. We just let's pretend we just received this NDA from Hitachi Corp. I've done all of the contract intake, and now it's time to review this contract. I can go ahead and check this out for edit. Let's see if I can move this. Sorry. It's opened in Word, but, I don't know if you guys can see it. Okay. Check it out for edit, and then I can open the screen's Microsoft Word add in. I can sign in using my Agiloft KB credentials, and that will authenticate me with my Agiloft KB as well as my dedicated, KB workspace. And here you can see, the screens word add in, allows you to do a few different things. You can select from screens that you have personally built or downloaded from the community. As you get more and more accustomed to screens, this list could grow. You could have a screen for each specific, kind of document you review or each each kind of, use case legal use case that you might have. And then we also have community screens here accessible right from the Word add in. So if you don't have a screen pre built, you can get an expert first opinion right from the community. So this is a non disclosure agreement. For the purposes of this demo, I'm actually going to look for a community screen. I'll come down to NDAs and you'll see that we have actually two different screens, by two different creators, for NDAs. I can read a little bit more about each of these screens, and if I think it looks good, I can go ahead and run that screen on the on the contract. So let's go ahead and do that. So screen runs, you know, they can vary based on the size of the screen, the amount of standards and questions in the screen, as well as the length of the document. When we're running a screen, what we're really doing is parsing all of the information in the contract, and we're using retrieval augmented generation to chunk this contract into relevant chunks. And so we do this on the screen side to make sure that we are only sending the LLM the most relevant information in order to get, you know, the maximum accuracy as possible and cut down on, LLM hallucination. So that screen just finished running. You can see I have that weighted average score up here of 67%. On our community screens, one of the cool thing we cool things we offer is, a market average or a market data. And you can see that this particular NDA from Chachki, performed on this screen worse than other NDAs that have been screened with this community screen. It performed worse than 92% of them. The average standard score across all of contracts screened with this community screen is 81, and there have been about 2,000 contracts screened. Again, this is this market data is only available on community screens, today. You can opt out of community data by not using community screens. And just for a little bit of extra data, we're not we're not storing any of the, you know, specific information other than the passes and fails, the rate of passes and fails of other contracts that have been screened with this community screen. Okay. So we got our results back. As you can see, all of the standards and questions are gonna be right here in line. You can filter based on the the risk level that we we identified earlier. But what we like to see a lot of peep what we do see a lot of people do is just toggle the fails so that they can start to address the the the failures in this document first. And we'll just hop to the first one. And when we click on this card, it'll highlight the relevant document text or source language that that, backs up this fail. And it will also give AI reasoning, justifying why the LLM thinks that this standard has failed. In this case, the AI reasoning is that the document defines Chkotchke LLC as a party along with its subsidiaries and affiliates, which violates the requirement that each party should be a single legal entity. So that's great. It it found a fail. It identified it in the document. Usually, you know, the next step would be how do I rectify this this fail? And, you know, different contract professionals might might know exactly what they wanna do to redline this, but one of the superpowers of screens is that we actually offer you a redline to bring this contract text into compliance with this standard. And we do that in a very in as low touch way as possible, so that we're not extensively marking up the document, before you you try and send it over to a counterparty. So let's look at this red line. In this example, it it simply crosses out for itself and its subsidiaries and affiliates. And we've also generated a comment that that will justify this red line to the counterparty. Again, you can just hop if if you want this comment to match your tone or intent a little bit better, you can edit it right in line. You can customize the red line by prompting the LLM additionally here. But if both of these look good, you can just go ahead and apply changes, and we'll red line the doc and insert that comment as you. So that's a quick, you know, a really quick demo of how a screen standard can identify non compliant contract text and bring it into compliance. Before we move on, I want to, I wanna show you the the results of a question. So we have this, agreement effective date. And the AI answers for this question is the effective date of the agreement is the date. It is accepted by the party identified below. I don't think this has a specific date in it, and so the LLM tried to tried to answer this to the best of its abilities. But you'll see all of our questions kind of will work in a very similar way where we'll try and parse the document and get you answers to your questions in a scalable way. One of the other cool things that we have within the Word add in is a one click red line. So if you've got a screen really dialed in and it's performing really accurately and you trust that that screen is is identifying issues as it should be, you can go ahead and red line the entire document, in order to, you know, apply all of the red lines at one time. For the purposes of this demo, I'm going to hop over to well, actually, let's let's actually hop over to Agiloft. And can you yes. You can see my screen. And we can actually see that those results that we just got from that screen have been sent back to our Agiloft KB and now live under screen results. So you can see, we have this, single entity requirement that we looked at earlier. As we saw, it failed. There's the definition. There's the reason it failed and a few and a little bit more metadata. When we then want to we redlined this, so let's go hop back into Microsoft Word and rerun the screen to see if our score goes up and if the, single entity requirement has been successfully addressed. So we'll select fetch and rerun the screen. This will, again, take a a couple a couple seconds because the the document has changed, so we wanna make sure we're catching any changes that have happened in the document. And one of the cool things as well is while the screen is running, we're actually we're actually sending the data to Agiloft in that process. So this is done running. We see the score has increased to 75. That single entity requirement is now a pass. So all good there. And we can hop back over to Agiloft and refresh the screen results to see that they've been updated with the latest results from screens. So let's look for that single entity requirement. Here it is. And now it is marked as a pass. So that is one of the major integrations we've done over the past few months. You can now check out a document from Agiloft, open it up in Word, opens the screens add in, run a screen, and all of the data that screens collects as a part of running that screen will be accessible within Agiloft. One other thing I wanted to quickly show is that while contract review and redlining is screens bread and butter, we do have a few, a a few additional functionalities that we call boost. And these are, you know, redlining or or, you know, document review functions that we think will help you out kind of adjacent to, adjacent to your screen. So you can see there are a few booths here. One like redraft, which allows you to highlight a clause, give the LLM some prompting, and we'll redraft that clause, according to that prompting. You can also highlight a clause and click make mutual, which will, redline the clause to to make the clause mutual. We have some we have this really cool one called, generate risk summary, which allows you to, generate a risk summary and prompt the LM right as an email to my boss, Greg. And prompt the LM for a summary of all of the failures, within the screen so that you can share those outside of screens to counterparties, to collaborators. And it's super easy to copy this to the clipboard, paste it in your email client, or or start over to prompt additionally. Okay. So that's been a very high level overview of our initial integration with screens. With all of our screens data flowing into Agiloft, we think that that kind of the sky's the limit on what you can do with this data, whether that's, keying off approvals based on a standard pass or fail, keying off approvals based on a screen's average score, or, you know, even, putting together dashboards on how your contracts have been, being screened. For example, in this this in this dashboard, we have companies with the lowest screen results. We can see here that Omniverse Digital has the lowest average score across all the screens we've run for that for that customer or for that counterparty. That may necessitate, you know, internal processes to to to reevaluate how we negotiate with Omniverse digital. You can also see, performance by screen, and and that can break up into the individual screens you have in your company. So it looks like our lowest performing screen is the SaaS savvy high value purchases. That may be, an impetus to go look at that screen and make sure that the requirements that you've solidified within that screen, are really kind of market standard market standard so that they, they pass more often in in contract review. So that's that's the end of my demo. I'm gonna be happy to answer any and all of your questions. But with that, I'll kick it back to Anne. Thanks, Nate, so much. That was awesome. We are extremely excited. So, I'm so happy to have both Nate and Jack here. We got a ton of great questions, and it'll be interesting to hear, especially to see how screens can maybe work with some of our other a o AI offerings, to answer some of these questions. So let's go through. Let's start with alrighty. Can we run screens to compare a third party DPA against my company's DPA? Can screens do a gap analysis? So while we can't, you know, do a a kind of typical gap analysis, as I showed you earlier, there is a functionality within screens to generate a screen, from your paper. And we we have this sweet screen generator that can take your contract. You can bracket around things that you want to test for, in in the third party paper, and we'll generate a screen that should show where, you know, the third party paper is noncompliant with the standards in your paper. But as for, like, just one to one contract to contract comparison, that's not that's not really Screen's bread and butter. We're all about building and operationalizing playbooks, AI powered playbooks, that you can use in an ongoing manner. So two follow-up questions based on that. We had the, could we take a community screen and save it and make it our own? And then also, is a screen a playbook? Yes. That's a great question. I had it in my talk track, but, it must have fallen out. A screen is a playbook. A screen is an AI powered playbook. We call them screens, but that is what it is at the end of the day. It's an AI powered playbook that that basically solidifies your in internal contracting knowledge, so that you can scale your contracting expertise throughout an organization. And then as for your second question, yes, you can take the community screen and build off of it however you like. So this is a community screen. Actually, let me hop to this other one. The services savvy is a community screen. It has a lot of standards that may or may not align with how you contract internally. But as soon as you download it to your account, that screen is yours, and you can hop into the configuration and edit it however you want. So this is a great way of building out your screens. It's a great starting point. You can also copy a given standard from a screen into a different screen, with a click of a button. So we like to think these community screens are awesome building building blocks for getting started with your own internal screens. Let's talk about the community data for a second. So community data, if the client wants to use screens and they choose a community screen, their information will not be shared out of their KB unless they get a pass fail, and then it's only the pass fail data that's being, sent back into the community. Is that correct? That's correct. We won't, we won't track any of, you know, the offending contract text or, you know, the standard itself. The only thing we're tracking is the rates of pass, fail, and the overall score of, you know, that screen when it's run on your your document. So the entire market dataset that we have is just is just how many times did this standard pass or fail, and what was the average score of the screen when it was run on documents? Awesome. How is the community how are the community screens, verified? How is the data information verified and shared, to how do you become a part of the community? Sure. So, our existing community has, you know, a lot of early adopters and creators who we have very close relationships with, and they work with our team to build out a, you know, a draft community screen for a given fact pattern or given contract type. And we iterate on that screen extensively to make sure that it it returns accurate results. And the way we do this, I didn't really show this in the demo, but the way we do this is through validation. So when you run a screen, and you're reviewing the results of that screen, we allow you to validate whether the LLM answered or returned an, a response that was accurate. So if the response is accurate, you can thumb give it a thumbs up, you know, and we'll track that average accuracy over time. So we've done this with our community screens partners, to make sure that their screens are super, super accurate, and we'll show that accuracy within the community when you read the description of a given screen. On the flip side, if you have a standard that is consistently performing inaccurately, you can give this a thumbs down, and we'll we'll similarly track that so that you know it's time to to maybe hop in and start refining that screen. Wonderful. Thank you so much, everyone, for all the great questions. We're gonna keep going through, here. Is there a desired average score? Is there a number that people should be looking for? Yeah. I think the the score you want is a hundred, ideally. You're building out these playbooks to, basically document your contracting preferences. And in in a perfect world, you want a % of those preferences to be met. Obviously, we know that that that's not gonna be the case all of the time, but, a hundred is the score that that you're kinda shooting for. Awesome. Can you add questions on the fly as you're going through and reviewing? This is not something that we support today. As I mentioned before, most of the screen building is accommodated through our web app. We've found that it's a much better experience working within the web app to build these screens. But it is something that we're investigating knowing that, you know, while operationalizing a screen, you may find something that you think needs to be added to your playbook, and it might be, you know, super easy to to just add that in line in the add in. But for today, all of the addition of standards and questions and tuning of a screen happens within the, screens web app. Awesome. Okay. Let's see here. Does the Word app work via SSO integration? Yes. It does. I just logged in with my account, but, you know, my token expired, so I'll sign in again. And you can see, you can sign in with your Agiloft account, and that supports however you sign in with Agiloft regularly. So you can log in via SAML or OAuth, and this should work exactly how it does with your, KB. Awesome. So what are the plans for the existing Agiloft Ford app? Yeah. I can take this one. So, as you might have noticed, if you've been using the redlining functionality in our existing Word app, there's some similarity here. Our plan is to take our negotiation features available in our existing Word app, and to map them over into screens to have a feature parody in in screens at least of everything you can do today in the Word app around contract negotiation. We haven't got any plans right now to remove any functionality from the existing merchant of Agiloft Word app, but there'll be a point where we have everything you can do in the Agiloft Word app in screens. So feel free to move over to move over to the screens app now to take advantage, of these exciting new features, or you can move over later once we've moved over features that you are using. Wonderful. Before I, transition back to the pricing and packaging questions, one more here. What's the quality of the information or what type of LLM is screens based on top of today? Yeah. So screens today is based off of, off the shelf LLMs. For the most part, the main workflows and screens are using, GPT four o. But we at, at screens are constantly testing using that validation feature that I showed you earlier to evaluate the accuracy of different LLMs like anthropic. And, you know, as LLMs continue to evolve, we'll be right there and can, you know, can and don't know that we will, but we have the ability to swap over to LMS if we see one vastly outperforming GPT four o going forward. Awesome. Fantastic. Okay. Let's transition to some of the questions on the pricing and packaging. Nate, if you would would not mind going back to our slides, and if you could go up to one more ahead, one more after, the one in between. Okay. This one? This one. Perfect. Thank you so much. Okay. So we're gonna focus on this. So a lot of the questions I've seen just to summarize again as far as the pricing and packaging changes go. So today, if you are currently an Agiloft customer with AI platform, you can use screens the second we log off this call. There's a little bit of configuration. Look in your email. You should have, access to the Wiki article that you need to access it, but it is there, and you can start using it within your workflow. So and I just wanna clarify again, with that. For our new pricing and packaging, so if you were to become an AgileFT customer tomorrow or next time your contract renewals, time period has come, or if you wanna proactively reach out to your account manager or customer success manager to get on the new packaging, what is included in that base model moving forward will be contract analysis, review and negotiation via screens, and then that Ask AI piece, previously known as Combo AI. So that is what is going to be included in the core functionality. Again, this will not just show up in anyone's KB at any time. An admin, will have to configure, the pieces and and introduce it where in the platform and where in the contracting process you want to use it. But those are the pieces that will be available to you. I hope that answers those questions. Just a reminder, we're going to have another webinar that dives a lot deeper into the packaging itself and what an end to end AI enhanced contracting workflow looks like. So please keep an eye out for that. Again, do not hesitate to contact your account manager and your customer success manager if you have questions in the meantime. They are ready to talk to you about it. They are excited, and, they can answer any of your questions. And then, finally, if you all are gonna be at clock next week in Vegas, Jack and I will both be there. Please do not hesitate to stop by the booth. We can take you through an AI enhanced demo. We can talk to you about any of the new things going on or or answer some of these questions, in person. So thank you again for joining us. If we didn't get specifically to your question, we will be following up via email, But we will talk to you all soon and catch you at the next one. Have a wonderful afternoon. Thank you.