Name: Ask in Logikcull
Uploaded: 2026-03-17T19:46:11.350Z
Duration: 33 min 4 s
Description: Ask in Logikcull

Transcript for "Ask in Logikcull": Alright. Welcome, everyone, and thanks for joining us today. I'm Robert Hilson of Logical, and I'm really excited about what we're sharing with you this afternoon. We're gonna introduce something called Ask, which is a new AI powered search, and data synthesis and analysis engine, that was pioneered and revealed and is now available, in in Logical. Before we dive in, I wanna introduce my colleague, Ash Patel. He's an attorney, product consultant, international man of intrigue. Ash is a senior sales engineer. I had to do it to Yash. He's a. c sales engineer, for Reveal and, has more kind of frontline conversations about us than than just about anybody in the company. So, we appreciate you joining us. Ash is always. a pleasure. Yep. Happy to be here. Happy to kinda go through ask. Really excited to show Logical user base kind of what we've integrated into the platform. So excited to show it. Right on. So here's what we're gonna try to tackle, and we're gonna do this quickly. I think we're trying to keep this to thirty minutes, and we do wanna get to as many questions as we can. So we're gonna start with how ask actually works under the hood, not technically so much, but functionally so you you understand what's happening when you actually ask a question and kinda what the purpose of it is. And then we'll look at kind of where it fits into your actual workflow, really from early case assessment, like, all the way through Depo Prep. And then we're gonna spend the bulk of our time actually demoing it. I'm gonna turn it over to Ash. He's gonna do a deep dive. During that time, I do wanna encourage everyone to ask questions. And, again, I think we have 250 or so people on this call. We'll get to as many as we we can. If we can't get to any in the course of the presentation, we'll give you our contact information, after the fact, and we'll, we'll try to get to you if we can. So, before we actually get into how this works, let's briefly touch on, like, what it is. So, ASK is a search and data analysis and and synthesis tool, that really allows you to do what what other LLMs allow you to do, which is pose natural language questions about your document, document sets, or entire document sets, and get back cited answers that help you understand kind of what's happening, in your matter. So if you need to come up to speed quickly on key players, if you need to build timelines, if you just need to kind of start to inform your search strategy, ask is is a great place to start. The core idea here is is pretty simple, which I think is what what makes it powerful. You ask a question and ask does really two things. It's gonna search your documents semantically, meaning it's looking for, you know, meaning and intent, not just kind of the the copy in the documents. And it surfaces the most relevant material and synthesizes it into, you know, in this case, clear answers. One of the critiques that we hear of some of the AI tools on the market, and and we tried to account for this when we were building really, when we build all of our our AI reveals, Adi works this way, and Ash works this way as as well. Like, we we wanna build it to actually show its work. We wanna be we wanna build it to be both pragmatic and transparent. Meaning, you know, we want you to know kind of how the the LLM is reaching its decision. So, critically, ask is built so that every answer comes with citations and links back to the actual documents that are in your document set. And kind of the net net of that is, like, you never have to trust the AI. Right? You can verify everything. So, when we get into, like, where to actually, use this thing and if I can go to the next slide here. There we go. What I find compelling about Ask, and and, Ash, I'll ask you to chime in here in a second, is that. it's not kind of a a point solution for one part of the process. It fits across the entire life cycle of a matter, at least from kind of that early case assessment, all all the way through, you know, trial prep. So on day one, you know, before you even review the doc, you can start actually asking questions and building case strategy based on, you know, what's what's actually in the documents. Ash, I'm I'm curious. You know, based on your customer conversations and demoing this, like, where are you seeing people actually getting the most use of this? Honestly, it's been all three facets, really. And I was gonna run through an example of each when we kinda show it live and logical right here. It's great as a tool of when you load data in and you are just searching or have natural language queries about what's in this dataset. You have as a practitioner, you have the motion, the complaint, whatever legal proceeding or filing in front of your face. Then you have the documents. And the way the genesis of search and kinda the legal industry has evolved is we know how to do our Boolean searches. We know how to get hits across documents. We have to then read, synthesize through thousands of documents to get the answers we want, where we can come in to ask, query up that very natural language question, and leverage it for either case strategy or review and investigations. Right? So it's been in all three facets that we've seen this leverage. And even one more at the end, which is a QC measure. Right? We go through our entire workflow, our process. We do our tagging. We do our redactions. Right? Then you drop in there, Hey, are these documents relevant to X, Y, and Z? Or if you know perhaps what opposing counsel's case strategy may be or theorizing what it may be, what in these documents that we're producing may be helpful for them in their investigations, right? So that QC measure's really coming up with ASK as well. Excellent. And I I know you're gonna show, show us that when we get into the demo. And in fact, we're gonna cover, at the end of this, some kind of, things to do with ask and things to not do with ask, you know, things it is, things it isn't. But, why don't we get right into the actual demonstration so people can actually, see what we're talking about? Absolutely. This screen share. is the moment of truth. That is call. I think you might have to stop sharing so I can. Alright. You know what? That would help. There we go. That would help. There we go. Alright. So, hopefully, everybody. can now see my window. Right? And, look, I know we have a lot of people in the audience. This is logical right here. And a quick refresher, I tend to say always in a lot of my either customer conversations or demos, deep dives. Right? If you've used any shopping website, anyone. Right? Macy's, Home Depot, Amazon. I'm not partial. I'm not sponsored by anybody, but you can get up to speed and running in logical search at the top. We pre filter the data for you. Your documents and your grid view, everything at the bottom, all of it's completely customizable. The newest, the latest, and the greatest is kind of ask inside our document dataset right here. And where it is is in the search toolbar right here. So that idea that even when we brought it over from our enterprise platform, the idea here is when we wanna query up our data, when we wanna ask questions of our data, we intuitively go to search. That's where we know where to do our Boolean searching right here. Right? Ask lives inside of the search toolbar right here. It's a part of the processing pipeline. And I'm gonna come in and type a very kinda simple question, but one with kind of nuances to it right here. Right? Who saw, oversaw finance at Enron? And this is a little bit of the best practices we're gonna get into in the second half of kind of this this presentation right here, but here's what ASK is doing. Right? I always say this, which is it's a closed loop system. I think that's important to know. It is searching only the documents that are in a project you put in right here. So right now, it's querying up about a half million documents right here as it generates the response. Even in this simplistic question, there's a few best practices I already put in here, meaning layering in A good example of asking a question for to ask is being narrow, having a scope with your question right here. If I simply typed in who oversaw Enron, maybe I'm getting back, you know, higher level people, people of the c suite nature, CEO types. Right? I'm asking for financial right here. I'm specifying at Enron. So within a dataset, adding just a few words right here to be more precise, to be narrow, to have a defined scope, to leverage ask in terms of how are we asking the questions. The reason I ask a pretty simple question is, you know, how often are we running through our datasets right here, our documents? And we are in a document, and we have a person, a query, a concept that comes up, an entity perhaps. And we just simply wanna stay in the search we're in, the batch we're in, the documents we're in. Right? Ask is in the search toolbar right here. Being able to kind of leverage something like ask to say, who is this or what is that or define this right here. Right? What do you get back when you query ask in kind of this chatbot interface right here? Two big things. Right? You're gonna see this narrative right here. Now, the narrative may have 15 citations. The narrative may have 10 citations. It may have two right here. It is going to search your dataset. And what it's gonna do, to Robbie's point earlier, is it's gonna come in and provide you citations. I think hallucinations become a huge part of kind of leveraging chat software or chatbot type. We anchor this to these 400,000, 1,000,000 documents right here where if I wanna click on the citation for where was this sentence kind of compiled, where was it figured out at, I can click it. It's gonna highlight it. I can put my eyes on this document. And for users of Logical, I could literally open up this document behind on my screen right here. Now, this is step one. The idea that we can come in, we can get and query ask with a question, we get a narrative back right here. Now, what we get on top of this is supporting documents. So we give you up to a 100 relevant documents to the query that you type in. One document may have more than one reference right here. Right? So the idea that one document can have multiple references, and we rank them kind of in where or how are they more likely than not to answer the query that you asked at the top right here. Right? So at any point, I can open up all 86 of these documents. They would automatically pop into my search back here, and I can start interrogating kind of further within the query or the subset here. I'll pause here, kinda check-in on questions. Robbie, anything to add right here? Yeah. Ash, we got a a couple questions. Yeah. And maybe just to to hit on the point, this is it's only searching your documents. Correct. Right? Be so you can you mentioned the hallucinations. Like, this is not pulling from Internet sources. It is just pulling the materials in your document set. So that's that's kind of one clarification. The questions I'll I'll hit you with them. back to back. First one is, does it does it search your whole current dataset and previously tagged documents? Yeah. So I think that, literally, my screen right now, right, I have the option in the sense of when you start it in a new session, it is gonna search the dataset or the deduped dataset right here. Now I can pinpoint this to a specific search. I can point it to the 42 I just added to my search right here, meaning I'd select this radio button, and I could start drilling in, get deeper and deeper. And my my advice with ask is always start broad, then get narrow. Right? Once we've tagged our documents, our relevant documents, our privileged documents right here, this is where we can pre filter or leverage the filters. I'll clear these documents out. Right? This is where I can come in and say, only wanna search my responsive documents right here. These 149, that's my scope that Ask would then search. I could come in and layer over 42 documents. So if I open ask again, let's say we had our tag filter right here, I can say, let's search these one forty nine for this next question I ask right here. So that's one way to kind of leverage pre filtered datasets or come in and build out a search and say, hey, these 2,000. I wanna search right here. How do I do that? What other questions? Good stuff. Kinda similar, but once you once you ask a question so say you ask a question on a set of tag documents that all fit kind of a common theme. Maybe they they might be issue tags, for instance. Is. asked then, is it assessing those and and and kind of learning from the results? Like, is it iterative? Yeah. So actually, as we roll out Ask into Logical right now, I think the feature is more about piggybacking kind of in the sense where it learns from the question. It is kind of doing it on an independent dataset right here. As we build out more AI capabilities with Ask, the intent analyzer behind Ask, right, we're gonna have more functionality in Ask as we get deeper and deeper with making it intricate into logical and embedded within logical right here. So today, the answer, no. But very soon, it should be. Awesome. And then there was a question around, is there a use case for Askit that can help with the tagging? Let's start there. The tagging and the review process, but let's start with is there is there ask functionality that just helps you with the tagging? Yeah. So here here's one example I have right here. This is a rather verbose kind of sentence. Right? And then we'll get into additional instructions in a second here. But from legal litigation, all realms of legal life right here, the question or the query that I pipe in right here is very much an RFP. Right? We've seen request for productions. The idea here is I'm gonna query the entire database on this. How could I leverage this with kind of coming in here and actually answering the questions from this document? Right? So let me clear up this question real quick. How could I leverage answering all my tags and you leveraging ask with a dataset right here? So what we can do is of course, it's gonna run right now. What we can do is let ask run for kind of, are there any conversations or documents related to the EES services right here? What could I do with my tagging right here? Right? I get my paragraph back right here and with citations again again. If I wanna see the fourth sentence, where did it come from? I can highlight that sentence. But I get a 100 supporting documents right here. My workflow very much would be if I'm starting a case, I'm taking all those RFPs. I'm taking questions I have that, you know, our our team has formulated right here. I'm popping open more documents, open these more documents. Right? And this is a two parter. Logical has suggested tags within it. Right? So there's a filter right here, which is suggested tags. It has some light AI capabilities within it, which is the more you tag, the more it's gonna learn. Machine learning right here. Or I can very much come in and say, hey. All of these documents, I wanna select them all in the search, and I wanna tag them responsive right here. So that's one way of attacking how would I embed this into my existing workflow, query up a question and ask, get the results, add them to your search, and then fire off the results within the documents and tag them that way and leverage our bulk kind of tagging to jump start some of those workflows as well as our suggested tags here. Right on. We just have, like, seven questions come in just as you were answering the last one. So we'll we'll maybe hit two more, and then we'll we'll let you keep brought in. There was a question about supporting different languages. Can you can you speak on that? Yeah. There are a couple different languages that are supported. I know Spanish, French are a couple of them. There's, I think, 10 languages that are supported right now as we've rolled it out since q one, but there are more coming in the pipeline. But there are several languages supported. Right on. And there was a question around is Ask AI, is it built on the concept of of clustering? And there was a separate question around which LLMs do we use. We. use a few different ones. We don't typically disclose those, but they're the ones that you kind of most hear about. And we have customized those to fit the the the data infrastructure of Logical and are adding new parameters to them to to kinda make this make this thing sing. Yeah. What other questions have we got? We got a ton, but why don't you why don't you hop back into it, or else we're. not gonna get. Absolutely. So look. One thing I wanted to say is, you know, go back to our workflows today. I say exist and logical. Right? We get data. We luckily can quickly drop it into logical. That's its highlight. Right? And then I can come in and tag the documents. As I come across concepts or queries, right, I could say explain, wrapped it right here. Now this part, I showed you the q and a nature. Right? Ask question. Get natural language answer. Get the narrative with additional documents that support it. Right here in the optional additional instructions is where you can really take control of the prompt or kind of how you want this narrative to come back. So I can very much say after answer, answer in timeline, explain the details, provide search terms. Right? And that's exactly what I'm gonna do where I can come in here and say, explain Raptor. And what it's gonna do now is do exactly what it's been doing in the last couple. It's gonna generate an answer. And after that, in the additional instructions, it's gonna say Boolean search terms that I can leverage. So this is kind of we talked about different verticals that's been used in. This is flipping it on its head. We know how to do search term searching. We know how to get the hits. But maybe this time, we leverage ask to actually get those search term hits right here. I may know a concept that I wanna further interrogate, query. I can leverage ask to actually create those search terms for me. Or to Robbie's point from earlier, if I wanna answer in timeline, very much at the bottom, I could do so. I could create a chronology right here, explain in detail, and create a list. So different ways to leverage additional instructions. And what I tend to say about this is I think in the past three to five years, all of us have all had to master AI in some realm capacity. Right? Prompting is kind of that additional piece where we can generate and kind of get specific with particular kinda questions and answers right here. So any other questions while we wait? There was a question about is there are there any document limitations, around there for best practices? Document limitations, no. It's gonna well, what you can do from a document limitations is what we kind of went over in terms of you can point to tags, documents that have been tagged. You can point to the entire dataset that you've put into a project right here. Different things in terms of limiting the pool of documents that BassSearch by adding them and then filtering on them and then pointing only to that with the radio buttons when you start the new session, but not limitations. And just quickly, wanna say, you know, this answer right here, explain. Raptor, you get your citations back. But right here at the bottom, Boolean search terms, you start seeing a couple right here. And this very much works right with Logical. Right? I can take this query right here. I can paste it in the back. I can search across my dataset. All of sudden, I got 640 hits for other documents maybe that I wanna look at leveraging ask in the dataset right here. There was a good question just kinda using your example of the Raptor project. Can you. ask for a list of projects that are referenced, or do you have to kinda know the project in advance? Now so with that, you could come in and, ask for other documents or other types of projects that are related to Raptor in this instance inside of this project right here. So definitely are able to do that. There was another question about the idea that we're going to surface the top 100 kind of similar documents. There was a comment to the effect of it would be helpful if we were able to do more than that. I can speak to the fact that I know that this is kind of a fast follow on, on our road map that we are kind of on building out that functionality. But is there anything you can you can speak to there, Ash? Yeah. No. I know that's a road map item. Right now, it is the top so what ask does, by the way, is it is gonna search the dataset right here. And the narrative component in the search results that come back are those relevant 100. It ranks them. It tries to find the most similar that are kind of related to the query that you ask. So it does search the dataset right here in terms of it's more expansive than kind of just a 100 results right here. But in terms of what's relevant to the narrative that it provides and the supporting documents it provides, that's where we kind of give you the most 100 kind of ranked relevant queries to the query that's entered here for ask. Awesome. So here's a fun one. Right? In terms of I'm gonna ask what is the EP project, but this is one I really like in terms of if I typed in EP project in quotes, right, I can come in and see that, hey. Do I have any hits for this at all? And I see zero. And this happens all the time in litigation. Right? We come in. We have a search term, and we got zero right here. I think this is one of the most fascinating use cases for Ask, which is how can I come in and leverage Ask to say that what is the EPE project right here? Right? And when I do that, again, the radio buttons is where I can limit the pool of documents. But as I type this in, it's gonna start searching the dataset to give me this answer back in terms of what is the EPE project, what is it related to, how come I perhaps didn't get any direct hits for the EP project when I searched it in quotes right here. And you can start seeing I think the question was asked earlier about, you know, different languages it can support, but you can see that, hey. This project, I know I'm reading for the text on the screen, but this project relates to something called the Kuaba project right here. Right? I have one citation right here. I have multiple kind of references inside of this one document, but I can open up these 29 different documents and start really interrogating how does this relate to a document or a search term that I got actually zero hits on right here? Right? Different ways you can come in and leverage Ask within a dataset right here. Other questions related or popping up in the chat right now? Yeah. How many do you? want? Hundreds of them? There's another question around, particularly for, you know, HR matters and investigations. How does the ask handle, or does it handle offensive language? Well, you might slightly whoever that is, teeing me up for so I just honed in on a little bit of a Slack data set. So this is a two parter right here. Right? So whoever asked us, thanks for this because this is kind of the segue into kind of how I'm gonna leverage ask next. But this is a two parter show. Like, look, Logical handles short message data. You get Teams. You get, in this instance, Slack. Right? We pre filter it. We can get you chat, conversations. You can hone in on reactions if you want to. But what I did right there, kind of to the earlier question asked, is just go into my Slack dataset right here. If I want to, I can come in and very much start a new session right here and say that, hey, I only wanna search these 21 documents. And I may wanna ask, are there improper relations? Right? So the idea here is look. Ask is great in the sense of what it can suss out, especially as an LLM in terms of the text that's used because it's using the text within the documents, it's gonna try to pick up, are there inferences to things? It's not gonna be always give you the brute force yes, no, but it's gonna give you, hey. Maybe there's suggestions of an improper relationship just in this chat message data right here. So this is where you can really leverage, you know, HR matters. How does it handle offensive language? It's gonna pick it up. Depends on the query that you ask it. What's it gonna do right here? How can we kind of come in and leverage it? So with this, you can kinda see that, yes, there are improper relations right here. You can start seeing a couple citations. And, one thing I haven't shown so far with this is I can see that this third sentence, it's talking about, you know, an improper thing, bad idea, something about a competitor. At any point, you can open up literally this one document for the citation. If I click this little button right here and it's directly embedded into the system where I can come in on the back and start reading, you know, the snippet of this Slack message right here in this twenty four hour chunk and leverage how do I use kind of that HR use case with ask in this sense. Pause there. Answer more questions. Yeah. That's great. Ash, there's a question around basically, like, what content is asked pulling off of pulling pulling from and specifically around, you know, is it pulling language from PDFs, from text in images, from handwriting? And I'm I'm gonna go out on a limb and say that that's it's probably the case that if Logical will index it, then it is it is capturable by ask. Is that the case? Bingo. And to that, we could get deeper in that answer too, but it is the document text. Right? And it is gonna when we're processing the data, there's a new auto tag leveraged and leveraged for ask specifically where has ask indexed the data and how many each of this data has it indexed. And you're gonna be able to flag kind of what ASK is what's there within your pool of documents that can be cleared up on ASK and perhaps some documents, perhaps logo files, right, that have no text, that aren't gonna be relevant to leveraging or querying ask or leveraging a question with ask on. Yeah. And and Logical will, you know, for the most part, barring exceptions, OCR handwriting, and it. is gonna capture, like, embedded text and images as well. So, you know, if if if and when it does, it's going to be capturable by by ask. Absolutely. Ash, there's a a question. And and by the way, we'll probably go, like, three or four more minutes here. And then there's a lot of questions around kinda how do I get access to this? What is the pricing? So. we'll we'll hit before we we leave. Can you explain further when the what what the intent analyzer you mentioned or when it will be developed and available? It. is look, we're on a webinar. So you know me. I have to try to be transparent. Roadmap's road map. But, look, we're we're the what the blanket statement I will put out is this is we have AI in Ask or in Logical in the sense of suggested tags, which is right here on the right. Right? We have brought kind of the next generation or generative AI with Ask in here. We wanted it in every step of the way with our data scientists. We don't want other people to be data scientists to leverage this tool. You have a question about your dataset. You have something you wanna ask or interrogate your dataset. You can very much ask the question in a natural language. For those of us that have become AI experts and have to check-in with AI every day of our lives, right, you can start doing additional prompting right here on the chatbot. There is more development coming with Ask as we progress through this year, the Ask Intent Analyzer, different things we're going to have with perhaps productions, right? But all of that to be developed and kind of released in a nice regular cadence here. Yeah. Excellent. There was another question about document size limitations, but but, again, the answer is no. I mean, I can basically if I would I can query my entire document set, without really any limitations. Is that the case? Yeah. Pretty much. I mean and look. You can can be large PDFs. Right? It's really gonna come down to how much text is inside of that document. And we can get more into the specifics for whoever has a question in terms of just reach out, contact us. We can get into the technicals in the weeds. Yeah. Fantastic. Ash, I think back to, back to you. Oh, I mean, I absolutely This is a lot of what I wanted to show right here. Now, there is a couple use cases. Here, I wanna actually hit on one thing. Right? I can show more of ask. I can show questions. I can show how we could do radio buttons. I wanted to pop this up one more time before we kind of show anything on our screen, which is, you know, when we're saying ask leverage ask, use it for case strategy right here, use it for early case assessment, review, investigations, couple success stories that we wanna share. Right? One third thing that we talked about is leveraging it against QC right here. Right? We do all our work. We tag. We can leverage ask to jump start our tagging. We can get our supported documents bulk tagging them just like I showed earlier. When you're all ready to go in terms of witness prep, dep prep, this additional instruction, you query those key documents. You query kind of the 10 and say, after answer, provide five questions, 10 questions for a deposition hearing. Witness prep right here. More importantly, perhaps, you know opposing counsel's case strategy. Perhaps in that relevant chunk of documents, you point ask instead of to all documents just to those relevant documents, and you pipe in what you may theorize opposing counsel's case strategy to be and be like, what documents are helpful to, you know, identify x, y, and z right here? So that's really where we've seen Ask Leverage. And also, kind of as a stopgap, right, in terms of sometimes people are leveraging Ask and saying, It's their QC check to whatever they tell their clients at the end of the day in terms of what they write their memorandum. No. We didn't find any evidence of X, Y, and Z, and they leveraged to ask to pipe the relevant dataset right here. Awesome. Actually and we'll probably take a couple more questions if if we have them, and then I do wanna get to, again, how to get access to that to this. There was a question around what document types does it support, and does it have built in OCR capabilities? And, again, I think this all goes back to, Yeah. you know, if it is if it is captured in the processing pipeline. and it is text, it is you know, it's going to be picked up by ask. 100%. And that's kind of why I show that example of the chat message data. But, hey, look. Logical itself, one of the greatest things it does is audio video files automatically transcribe them. That's transcribed text that lives in ASK's index. We We can search against that with ASK. Awesome. Alright. Ash, if you can let me share, we'll go back up. to Absolutely. how to get alright. How to get access to this thing. So we'll make this available to you, kinda irrespective of of where you are, you know, with respect to just trying out Logical for the first time or subscription customer. I'll try to briefly walk through these. So, if you if you really just wanna try this thing out, and this is the case with just Logical at large, we do offer it in pay as you go plans. It's basically a small premium to what your normal pay as you go rate would be. So it's just, an additional percentage of the the monthly storage cost. Pay as you go is month to month. If there's no commit, we don't lock you in. And, basically, the way it works is you're gonna get, you know, a number of questions that you can use per gigabyte. If, you know, you're looking for kind of, more predictability and use, if you're investigating logical for, you know, that matter velocity that you know that you will have, a subscription option is is probably the best plan there. And, again, the way that we actually price this thing is we're gonna give you, you know, a block of questions per gigabyte and then just a small premium to that per gigabyte rate. And then this is important. If you are an existing subscription customer, we're offering with no strings attached the ability to use this on one matter, for free. And so, if you wanna use it on an existing matter again, if you're a subscription customer an existing subscription customer, you wanna use this on an existing matter, you can basically use it on anything that is below, or at 250 gigabytes. And what we're gonna give you is 250 questions or or prompts to use. For net new matters, the only limitation is it has to be under 500 gigabytes, and we're gonna give you 500 questions for the for that matter as well. So if you're interested in learning more, the best way to, you know, to get in touch with us, one, can just go to our website, you know, ask for a demo and reach out and say, you know, want I'm interested in Ask. We'll point you to the right people. You can also reach out to either me or Ash. We'll make sure that you get in contact with your customer success team or your account manager. If you already have, you know, a a contact internally, your account manager, your customer success team, your sales rep, reach out to them. You know, they're they're they're all aware that this is now available, and they can, they can get you set up. So, Ash, with that, man, that was thirty minutes went fast. That did. Anything from you to close? No. Look. I always say this. I think you put a perfect bow on it. Anybody that wants to try this out, anybody that read that last site, get in touch with either of us. Get in touch with anybody, your account executives at Logical Reveal. And, look, appreciate everybody's time for hopping on and checking out the product. Yeah. We we really appreciate everybody's time. The last thing I'll say, and this is, you know, a comment on several of the kind of road map related questions that we got. You heard our our CEO say this at legal week many times in many different outlets last week, but, like, Logical and the other products that are under the Reveal umbrella, you will see an increased pace of innovation and investment around these products. And so, you know, this is this is the introductory ask. We have a bunch of fast follows. In ask, we got a bunch of great FOIA stuff coming down the pipe, automated redactions. I mean, it it it's really exciting stuff. And, Ashley, like, the the, you know, the the resident in house, logical fanboy, like, I'm I'm clearly pretty excited about it. So, I appreciate you. Thanks for being here, to the hundreds of people who joined us. We really appreciate your time. And, again, reach out if we can, kinda help you put, be put in touch with the right person. See you all next time. File.