Name: Unified Written Discovery: Identifying Discovery Responses with ASK in Logikcull
Uploaded: 2026-04-22T20:16:06.654Z
Duration: 51 min 24 s
Description: Unified Written Discovery: Identifying Discovery Responses with ASK in Logikcull

Transcript for "Unified Written Discovery: Identifying Discovery Responses with ASK in Logikcull": Hello. Hello, everyone. Hello. Welcome. Welcome to today's webinar. I'm Jacqueline, and I'm gonna be walking before we jump into our exciting topic today, specifically about, on about unified written discovery, specifically going a little bit into asks features within Logikcull. I wanted to give a couple of high level quick notes to people. I get this every single time we have a webinar, so I just it's great to be upfront. Right? This is a recorded webinar and will be available on demand using the same link that you got in your email, and we will this is live, which means that you can go ahead in that chat or q and a feature on the right side of your screen. Go ahead and jump in and ask us questions. I'll be monitoring the chat throughout the webinar to make sure if there's anyone having any issues I can support in any way, and And also, I turn the back end that we're answering questions, and I'll prompt our speakers. And we'll have a couple of different polls throughout as well as survey at the end. I love to get your guys' feedback so that we can make even better webinars in the future. So without further ado, I'm gonna pass it off to my friend Josh, and we're gonna jump into the topic a little bit. Hello, everyone. My name is Joshua Gilliland. I'm a California attorney, and I'm here at Reveal helping with content marketing, getting to put together webinars like this one. With me today is Ash Patel, and we're gonna talk about ask in Logical. And we're going to have an approach that I feel is very important for any attorneys out there of thinking of written discovery as a unified front because you have request for production, you have special interrogatories, and someone can't hear. So, JQ, can you hear us? Okay. We'll figure out what's going on. JQ, if you could give them a hand, that'd be fantastic. And we're gonna talk about written discovery. So if you're online watching live, please say hi in chat. Let us know if you're an attorney, if you're a paralegal, if you're an ediscovery professional, trial presentations specialist. Let us know who you are and where you are from. So today, we're gonna talk about features and ask with some mock written discovery. And the way that we're going to do that is not with the Enron data. For the last nineteen years, a lot of product demos have been with the Enron data. I'm a little tired of that. We have all of the FOIA data that the government has produced about UFOs. So we're gonna have some fun with the UFO records that are real, that, yes, the truth is out there, and we're gonna play with it and do some searches in it. And for those who, enjoy science fiction and this topic, we have a survey for you. What is your favorite UFO movie? Now there's some qualifiers for this. We don't say favorite alien invasion movie. We didn't say favorite alien attacks movie, because there are a lot of genres here. So we're we're keeping it more in the flying saucer sense. So go ahead, answer the survey as it comes in because we're curious, like, do you like the day the earth stood still? Do you like arrival? There's a lot here, and there's no way to capture all of them. But, again, just to have a little fun with this subject matter. And judging from the chat, we have some very playful people, and that's fun. So, Ash, could you tell our audience a little bit about yourself? Yeah. Absolutely. So my name is Ash Patel. I am a senior sales engineer here at Reveal. In a former life, I was also a project manager handling kind of these ediscovery requests all the way from getting the search terms done, getting the interrogatories, reviewing the data, compiling the search terms to find those, you know, documents that may be responsive all the way through production. And then in a former life prior to that, I was also managing one of my own practices with a colleague as well. So that is a little bit about my background. And my answer to the question, what's your favorite UFO movie? I said it yesterday. I changed my answer as I thought about it. It's still Arrival. It's Arrival as an adult or older adult. I have liked Arrival, what it kind of says about if UFOs arrived, how do we handle them. But, you know, as a kid, younger, it might have been Signs. It might have been ET. Goes between those two in terms of which movie may be your. favorite as towards, you progress through life. Close Encounters of the Third Kind being. someone who was born in the seventies. I know. I I'm a big Steven Spielberg fan, and I do wonder if disclosure day is actually going to be a secret sequel to Close Encounters. So but, you know, you have JI and helmet Hynek has a cameo in it and, you know, flight 19 and a bunch of other fun stuff. And we're gonna get to play with that. So, Ash, let's imagine that you and I form Patel in Gilleland, and we start handling discovery, because, you know, we we have the litigation bug and we wanna help people again. And not that we aren't now, but let's just say that we're representing people. You know, we get discovery. And I like to think of written discovery unified because something that always gave me stress was, hey. I'm gonna work on the request for production and then work on the interrogatories and go like, oh my word. They're related. We should have attacked it together. So at my old firm and and other times, I've tried approaching request for production and special interrogatories, like coming up with a a key code to keep them together. So I know that RFP two and ROG two go together. In real life, it could be RFP two and ROG seven go together. And, you know, that that's kind of exciting on on how to handle that. How did you handle matters like this? Yeah. No. Absolutely. It really came down to looking at the entire set of written discovery, either holistically to figure out perhaps, you know, some RFPs or interrogatories that may go together if we could loop them together, especially as we, you know, progress and talk about issue tagging a little bit later on. And then it also came down to what are they requesting in terms of as a specific custodian individuals that you're looking for? Is it topics or themes that we could categorically put together to better help craft these searches we'd go through in the data that we've collected either from our clients or from opposing parties, whatever it may be. But it'd always be taking an overall look at all of them and then perhaps bucketing down further than that. I like that approach, and so we would Jill will, if we were at a firm together. I would also add for for everyone, when you're gonna start document review, set up issue codes to begin with. You know, think about your search strategy and, like, have that approach because I don't wanna have to repeat document review if at all possible. There might be a time you have to double back because you're gonna learn more things as you do review. Like, that's life that happens. But I won't big fan of issue code in advance. And just to highlight that, here's some sample issue codes, that are set up in in logical. So I have folders with the different RFPs or special interrogatories and request for admissions. We have some preset tags as well. And then in each folder, we have RFP one, RFP two, RFP three. And in doing review, you could go, hey, tag accordingly, or if something's privileged, tag accordingly. So my question for the audience in our next survey is, how do you use issue coding? There isn't a wrong answer to this. It's a lot of, you know, preference on how you attack it, But, let's let's take a look at what people, do here. Do you tag for RFPs? Do you, you know, rugs, you know, privileges, tags for causes of action? There's a lot of approaches here. No defenses, thinking about admissibility. All of those things can keep you out of trouble later in the case. And, yeah, Ash, how did how did you like approaching this? Yeah. Look. For issue tagging specifically, I always traditionally like you know, when it came to issue tags early on in the case assessment, I always found them very useful in terms of bucketing documents to help identify relevant documents that you may actually eventually produce or be flagging for some particular use case. It's not always production. It may be internal that you are bucketing documents in a sense. I always viewed issue tags as buckets, but you're doing it for a higher purpose right here. And with that, it has its effects downstream, whether it be production where you could respond to these RFPs, whether you could hit a tab just like you've organized in that slide right there and say, yep. All these documents are responsive to this. I am ready to do production one related to all of these obligations right here, where I may be doing a rolling production and still running my searches and documents related to the subsequent set of written interrogatories right here. But it's always identifying them, bucketing them just like you have, and then separating them out, below them, or nesting them in a sense right there as well. Cool. And, let's take a look at some of the results coming in. And it looks like, we have a bunch for 34% for tagging for request for production, 9% at rugs, privilege, 34%, causes of action, nine. People are still voting, so these are shifting. But, you know, there are definitely, request for production and privilege are at the top, and that makes sense. I would encourage people to think about interrogatories, as well because that could reduce stress, on responding to to those questions. Well, let's take a look at the fictional RFPs that Patel and Gilliland are gonna be responding to. And we have RFP two, which is produce all records relating to Kenneth Arnold's sighting of identified flying objects from 1947, interrogatory to identify all records referencing Kenneth Arnold in project saucer, and admit that project saucer investigated Kenneth Arnold sightings of saucer like craft. So, again, we've put these together logically, and now we can approach. So doing the, you know, baking show method, you know, I ran this earlier, you know, put together a, advanced search and logical. And so with the name Kenneth Arnold and the word saucer or an unidentified flying object, and I come back with one hit. And that one hit, you know, checks all those boxes. So I would tag this for, you know, RFP two, ROG two, included for the RFA so I can respond to that. But let's talk about how we would attack that with ask. And I wanna stop share and turn it over to Ash so we could show how would we bake this cake? How do we approach this? Absolutely. So as my screen starts sharing right here, what I'm sharing is the platform logical right here. Right? And just a couple things about logical, which is, you know, you can come up, and those searches that you just curated are advanced searches where you can start clicking this button right here. If you wanted to do a basic keyword search, you could write it into the platform. You have your filters that really allow you to start queuing in on items and then, obviously, your documents to start reviewing right here. How I would actually attack the search is leveraging a tool that's easily available in Logical within it and right within the search toolbar called Ask right here. Right? Normally, we just went over this in that slide deck. This exact interrogatory or rogue will require multiple keyword searches, reviewer time, cross referencing, authorship with Kenneth Arnold right here, and then figuring out conclusions that may be relevant, may not be. What I'm opening up here is Ask. Right? And what Ask is is a generative AI tool grounded in actually your data, giving you a clear written answer tied to the discovery in your project. Behind the scenes, there's a lot of technology within Ask. It's using, you know, semantic searching across your datasets. When it gives you results, it's ranking the most relevant ones higher right here. It's using a little bit of rag behind the scenes. Now there's really nice things that ask does, but what I would do out of the gate, if I had this interrogatory is I would put it into ask and see the results that I get back. Now we're gonna click enter right here, and what this is gonna help me do is it's gonna identify documents in this database, especially specifically this project right here, and out of these 9,000 documents to my left. And the point to show there or share there is this is a closed loop system. Right? It's only going off your dataset and not from the Internet right here. So this is key. This is defensible. This is accuracy right here. There's a couple things that asks gives you back here in terms of beyond just a narrative or an answer to the question right here. Right? So with it, we can see that, hey. We have this narrative that comes back, and within the narrative, we have a couple citations that come back right here. So all these little footnotes, and I think this is great for attorneys, paralegals, anybody in the legal sphere, right, is anybody that's leveraged AI because AI has been the biggest buzzword in the industry in the last couple years. Right? We've seen things in terms of, is this accurate? Is this not accurate? Ask is very much gonna find the answer in your data if it's there, and then it's actually gonna come back. If if I ask it a question and it's not in the dataset, it'll say, based on the results, I can't find information relevant to that answer. So part one is it's gonna help try to find the answer if it's there, and part two is when it does find this narrative, it's gonna give you back these key citations. Now you may have multiple citations or what we call references within one document right here. And as this narrative is being built up right here, right, I can come in and say, hey. Where'd you find this third citation from? And it's literally gonna key in on that document and highlight it within that document for me on this summary page right here. So this is nice, you know, in terms of I can come into this document. If I wanted to click into a different citation, I can click into a different citation, really get a quick overview of not only this narrative with my query right here, but then I can also come in and start digging into exactly the site where I'm trying to figure out or glean that information from right here. Now at any point, I can open up one of these documents in the background if I wanted to. And when I do that, I have the ability to open up both of them. And one more thing that you get back with the ask is we give you answers that make up the narrative. So the top documents I know I'm scrolling here, but the top documents that make up a narrative, it could have two citations. It could have 15 citations. However many documents that we use to make up this narrative, we're gonna cite to first and provide up top. But in addition, we're gonna provide up to a 100 results that may be related or relevant to the query or question you asked right here. So we can see nine supported documents for this very targeted and nuanced kind of interrogatory that we asked right here. Behind this is actually the document right here. So one thing I wanna share is specifically with, you know, one thing that we're doing here is we are finally not showing you the Enron dataset. Right? We are not showing you King Pan and Vince Kaminski right here. We are actually showing you this UFO dataset, publicly available information, but this is scanned material right here. And I wanna highlight that. Right? Ask is taking this scanned material right here, letters, numbers, you know, and actually being able to parse through that through Logical's processing engine. And, specifically, we are looking for one person, kind of saucers within that person's identifying or interrogatory right here, but it's also gonna highlight right here. So the one quick thing I did is I click search. I'm looking for Arnold. I see that I have a 406 page kind of PDF right here. But real quickly, in a scanned PDF at that, I'm able to kind of pinpoint where Arnold is and start now reading and reviewing. And I kind of take it back to you here, Josh, to kind of highlight what Ask is doing and what we're reviewing in terms of Arnold and what he's saying. So the beauty of this feature is the fact that you get metadata like qualities from the scan record. And the context that's coming back is not the type of thing you would have had without generative AI to help create a summary of it. That is a game changer in my opinion because having worked on cases with old scanned records, whether it was insurance, you know, agreements or I haven't worked on one of these, but friends have toxic tort cases where you have scanned records going back fifty years. Like, that doesn't have metadata. And so you don't get to do the cool searches for emails from, you know, person x regarding subject y. Like, you don't have the ability to do those type of metadata searches, but this helps level the playing field if you have scanned records. Or in this case, some of this has scanned microfiche, and I've joked how often do you see microfiche in civil litigation and because you just don't. But here, the sample has records from microfiche that have been turned into PDFs. That's wild. So the idea of being able to get metadata like qualities from doing a search with ask, I think is very helpful, and you get similar results with metadata as well. So, again, I think it's a game changer if this is going to come up in your practice, because their paper still exists. It's not the big thing. People are writing emails left and right. People aren't, you know, conducting business like it's 1985. But the idea that you're gonna have scanned records still is something you should not, you know, discount. The other factor is the ethical rules that come in with GenAI that generally deal with legal research. And that says you should trust but verify. You should still go read the document. And some of the standing orders that are coming up in, state and federal, courthouses are you attest that, like, yeah. I've actually read the results. It's not just the summary that I looked at. I actually looked at the document because you don't want to be one of those lawyers who makes a giant epic mistake and then writes an apology letter to the court. Like, we have those happening multiple times a week. So you can comply with ethical obligations by being able to go, here's the footnote from the summary. I'm gonna go read it now. And that way, I am following all of my ethical obligations, as an attorney and making true statements to the court, with with whatever hits that you have based upon searches that you've done. So, so we're getting questions. I, I would on. add to that real quickly, which is, you know, right here what we did is we asked a question or query to ask. When I popped open this document, I searched for Arnold. I knew that was key individual that we were looking for. But right here is that part where this is our review workflow that hasn't changed. Right? What ask really allows you to do is accelerate that path to identification, review surgically at this page because now I'm gonna come in and say, hey. I wanna tag one of these documents, which I can start doing to the responsive RFP or interrogatory that was this was a part of. Then I can even go further and pin site this or annotate this if I wanted to with logical's annotation features. So really coming in, and instead of hours of searching and validating, we got a pretty defensible start right in a couple seconds to then go expand from this view further. Yeah. Absolutely. Well said, because it's never one and done. Like, you have to look at it. You have to go, does this actually go to the this RFP? Does it go to the second cause of action? You have to, you know, still lawyer, but it does give you a jump start. So we we have a couple questions that have come in. And if you could turn off screen share, let's let's answer them, and then we'll jump back into. the material. So one of the questions is it's a civ pro nerd question. And it's like, why would you ever answer an interrogatory asking you to identify all records instead of just producing them? This is if they ask you to identify in at least in California and in federal courts that I've been in, you do have to produce a list saying, like, here's all the documents that support our second cause of action. You don't have to summarize. Like, that's a compilation. They can go read it themselves. So but you do have to identify records. And so that's that's a, again, fine point and sit pro. A related question is if you put in an actual document request, will it find all responsive documents? I would not hang my hat on all. I would hang it on. It's a great starting point. And you can use ask to identify common terms with the hits that you have. You can refine the search, but it it's not it's a giant leap forward, but nothing's a magic wand that says pressed over done. I would still validate. And while, you know, you get into the nuances of rule 26 in federal court about, you know, perfection is not the standard. You know, you have to make a good faith effort. You know, you wanna avoid an inadequate production, but it's a great starting point. And then you, you know, test and validate to say, do we have everything? Because the court will understand if you miss two records out of 2,000,000. Like like, again, perfection's not the standard here. Now, Ash, do you have anything to to add to those questions? Yeah. I do totally agree with your first part, in terms of identify and summarize. Also difference in the obligation or responsibility to identify and then produce. Right? Two different things. They go through a process of identifying and then the actual process of going through relevancy and producing. What you produce may not be the original set of what you identify from and that changes state to state, which you're correctly identifying too. And then I think you've summed it up perfectly in terms of ask is going to be a great starting point. It is not the magic wand for kind of all. Because within all, especially within some of the questions that we're typing up right here, in any case people have had, right, we discover terms. We discover nexus terms, things that aren't directly tied to the key terminology or core issue at hand, but that have been used as a project name, a code name, different product name. Right? And being able to learn or glean that information through leveraging, perhaps, ask to identify the initial review set, then taking that to expand on, you know, identifying and validating with your eyes the other documents that may be relevant or responsive to other information you find for ROG right there. Well said. A follow-up question is what's the advantage of using ask over just searching the name? Just searching the name might be sensory overload if you're trying to have some nuance about one person and what they saw or what they did in a specific case. It's like with our hypothetical, you know, reference to Kenneth Arnold and Project Saucer. Like, if he's in there a bunch of times, it could take a while to find project saucer, as a related topic. However, just searching his you know, like, that one search that we did at the get go found just one record. You know, the search that we did with Ask found a lot more with more contacts, for us to go through. So, again, it's it's always an iterative process, but I think this expedites the, approach on being able to find and get insight, into your records. So, but, you know, good questions. And let's just make sure that we haven't left any oh, one other is, confirm that, how is ask available, to be on demand as in nondescription. JQ answered that and ask is available for purchase to all logical accounts and can be turned on by your account manager. So and we'll we'll definitely talk more about that, as we move on. Well, let's, let's jump in and have a little more fun. So we have another set of hypothetical, responses. And so we have a interrogatory for identify all records written by doctor j Allen Heinick that concluded that UFO sightings were astronomical in nature. And then we have two related, requests for admissions. Admit that doctor Hynek drafted reports that concluded UFO sightings were astronomical in nature and admit that doctor Hynek drafted reports that did not conclude you have sightings for astronomical in nature. And you can get diametrically opposed RFAs because not everything's absolute. And so as a practice point, you know, you can debate, do I deny part of this and put in an explanation? That can happen, or you could just say, admit. Yeah. That's true. He did write a bunch. So again but those are practice points. So, Ash, let's attack this. I'm gonna stop screen share, and let's get into, reports that doctor Hynek wrote and how we would approach this. Absolutely. So one thing I wanna share right here is, if you think about this exact request that we have, which is, all the reports that doctor Heineck had that were astronomical right here, I wanna go through that actual exercise that you did with the first interrogatory right here just to show a little bit of what logical can do. So we had the advanced search that you built out, but this interrogatory in particular has three sets right here. And where I think to the question that was just asked, why don't you just search for the name perhaps? Why won't you just search for the word? And the word here is actually a pretty vague, expansive search, especially when you're thinking about space or the universe or UFOs particularly. Right? If I have Heineck and I have the word astronomical, I'm doing this search perhaps basically this way. And this is to show that, hey. It still can be done in Logical the way you want to, and you can test out your searches too. This is one thing I've always really, really liked about Logical is if I test my search, not only can I get back, hey? Alright. Here are a couple things that I grouped together that perhaps all of us know in terms of astronomical high neck. This is what I'm looking for, UFO sites. Right? But that word astronomical, I'm probably hitting up the source or googling words similar to astronomical that then I can make sure is expansive and capturing all the results that I want going back to the earlier question of making sure we put eyes on documents to make sure we answer the question for that all segment right here. But just out of the gate, we see even five four eighty nine documents, but I know I'm about to review 500 pages to find this very targeted kind of ROG or RFP right here. So with this, you know, the point I'm trying to make is we are probably gonna have ask come in, and we're trying to say that we wanted to interpret this. We wanted to intuit the meaning here. And so I'm gonna open ask back up, and we're gonna pop this question back in here. And, again, with ask, we're not just asking it to do that keyword search where the keyword search finds the brute force search of all hits related to kind of that word that you put in. Ask here is actually interpreting a meaning. It's not just matching up words, but it's helping me find the documents that are relevant to the query or question that's asked. Ask takes a little bit of time, but right now it's popping up right here saying, hey. We have the a result right here based on the documents provided. Right? And this is a key thing that I wanted to call out earlier is it's almost always gonna say something like based on the context provided, based on, you know, the documents provided right here. Every result is traceable back to the source documents. So going back to the point we started with, you can validate. This is defensible. You can then tag and produce with however your ediscovery workflow may traditionally work out, but you can leverage us to accelerate that part into the review right here. And with this, we see a couple, you know, questions come back in terms of percentages right here. I could click another footnote right here just to see where this may be. And, again, I can open up this document if I wanted to. This minimizes my window right here, and this is gonna load a large kinda scanned in PDF right here. So this is where I'm coming in and actually seeing, hey. This is a large scanned PDF, another one right here, and I can look for I think it's Heinick right here's name, and I can search the document in this sense. Give me a second here. Of course. This is always related. We'll click this. Hi, Nick. Either way, Joshua, there's part you wanna discuss about this document while I get the search working right here. So it's, because we're looking at context for, like, what did he write and when. Because if you're not a nerd on the subject, and that's okay if you're not. You know, Hynek was part of project Blue Book and being able to come in as the skeptic to say, hey. It was swamp gas. And by the end of it, he believed there was something there. And so he does a flip over the course of his life. So being able to go, what did he write and when, for reports about, you know, 30% or some percentage were definitely, you know, Venus in the night sky. First, the times he said, we don't know what this was. And being able to get the context from these old reports, is is very handy and being able to jump into it as well. Now we do have another question. Can can this create a timeline? The answer is yes. And we'll we'll talk more about that because there's some other, functionality that's helpful for coming up with meet and confers and and other prep. So, but being able to go like, hey. The percentage of reports that he wrote, will help us answer the request for admissions and the special interrogatory. So, and also issue code to our heart's content on what this is responsive to. Absolutely. So, but, yeah, good questions, coming up on this. So, on that note, why don't we show the ability to create a timeline, since that's, been. asked? Absolutely. So let me share my screen again. I apologize. One thing I wanna highlight here while we're on this actually, while we're on, you know, this particular question right here is that ability to to add these four supporting documents to your search and really leverage ask to pinpoint your results right here. So there's two things I wanna show. I wanna show the answer or create a timeline right here. But first, let's take this Nuance kind of workflow we're in. We're in an interrogatory. We're looking for a specific thing related to percentages that he may have seen UFOs right here. And if I open up these supporting four documents, they go right in my search in the background right here. This is where I always say start broad with ask. Right? And then we get very surgical and narrow with it right here. And instead of me and tuning or reading everything, I could come in over the top and say, hey. Let's summarize these key findings in Hynek's report right here. And we could see if ask gets a result back or what it gets as a result back right here sometimes. Right? We're actually pinpointing this now to just those four results right here. And what we get is what Joshua just read out in that scanned document right here. Right? We can see that he's actually getting an additional 37% right here. We can see 30% right here. And at any point, again, the same thing I've been showing, we can see a footnote. We can see that, hey, Ask is building its results off of not just the document I've been showing right here, but also this TIF right here. And if I wanted to open that up, this is that idea that I can pinpoint this page right to this result where there's a final report. There's discussion with the consulting agencies, and this is where he's seeing extraterrestrial or astronomical and origin, you know, findings right here. So that's one way to really leverage ask is you come in, you have a subset, you have, you know, your relevant documents, your key documents already identified, you can pinpoint it just to those subsets of documents. And to take a step back before we show the timeline feature right here, all of these filters are the quick way to build out that exact same search that I did. Right? I could key off of just these 15 documents by clicking this tag right here. If I'm interested in specific document types, I have the ability to do that. People places, I can pinpoint on just Bob Smith's production or his custodial file right here, these 3,000 documents. Document date, that's where I come in and either manually input a date or use that calendar function we're all used to, or I can say, hey. I'm looking for documents from this to date, this date to that date, and apply that range right here. Now with kind of the answer and timeline question right here, one thing we've been doing this entire time is these interrogatories right here, right, which is pretty targeted right here in terms of what am I finding, who are these results, who are these people that I'm after right here, what do they see right here. But if we take a step back, everybody's kind of been asking for who saw UFO sightings right here. So I'm just gonna type that in. And this part right here is where you can actually take a little control of that narrative that you get back because you're gonna get a summary back here still. But this time, I'm gonna say, answer in timeline for me. Because I want a little bit of that chronology of whatever it finds in terms of UFO sightings. This is us taking not the, perhaps, 1,000 foot view right here, but this is us taking a 10,000, 30,000 foot view of these 9,000 dots related to UFO sightings right here. And what ask is gonna do is a two parter right here. It's still going to answer the first question I asked it, who saw UFO sightings right here. But in part two, it's actually gonna take an additional instruction right here and say, give me that answer in a time line fashion right here. And it's gonna create kind of that exact replica. So, again, what we see here is a nice six kind of cited narrative right here. I could click the six citation. I've been shown this in terms of who are those people. But right here at the bottom is actually that additional instruction in action right here. And when it answers in a time line, it can do it in a variety of way. The more descript you get and this is where we go back to the better your prompts are, the better your directions are for these machines or these robots or LLMs, the better the output is sometimes. Right? So I set answer in a time line, and what it did is broke it down into decades right here. In the forties and fifties, it tells me military and, various airmen found things. But then you really start getting into people, places, radio observers right here. Right? Different kind of people or, groups of people seeing when these UFOs reported right here, and it always provides those 11 supporting documents that then we can really go out and start building our answers based off the results right here. So this is kind of a good idea of the time line that it can build, and it can build more descript ones with a better instruction in terms of answering time line and provide detail about who saw x y z. Thank you for showing that. One of the other features that can be leveraged with ASK is identifying common terms, which you could use in a meet and confer if you are gonna talk about search terms. I think ASK can help you, you know, redefine that meet and confer process and being able to go, what are we trying to find here? Okay. You wanna find information about UFO sightings in the nineteen nineties. Okay. What are the common terms that are, you know, that you wanna go look for? Because it's easy to articulate what we want to find, figuring out the keywords to do so is always the challenge. And that's why keyword disputes get complicated and messy, and courts frequently just order people to go meet and confer about it and figure out what are the magic words you need in this case. Well, ASK can help with that process because there's so much guessing otherwise, and this can help, focus the work, you know, and again, nothing's ever a 100%, but this is a far superior starting point than guessing what might be the magic words you need, in order to find responsive data. So, and on that, we have another survey. So about, like, how do you like to do your searches? Do you like to do it based upon the written discovery like RFPs? Do you like to run searches based upon agreed upon search terms? Now we have that poll under the poll, you know, response and, you know, click the one that you do the most of. Because, again, I'm curious on how you work. Like, because by being able to answer those questions, we're able to help you better in understanding how you approach discovery. So, I'm in the RFP or special interrogatory club. I have many friends that like to get everyone to agree upon search terms. It's fancy taste or preference to invoke contracts, but, that's how I roll. Ash, how about you? Do you prefer the RFP exact? same. Okay. It'd be that maybe we should start this hypothetical law firm. It'd be the exact same. Getting the parties to agree on the search terms so we know what we're doing from the onset, and everybody's aligned in terms of what we're looking for and what our obligations to produce are. And thereafter, you know, having the duty and obligation to sift through your documents, review, tag, and produce said Awesome. Yes. We gel well together. So as. we again, if you have questions, please please, ask away, and that's not meant to be a pun. But we do have no time as as we keep looking at this. Why don't we, stop share, and we might come back. to that. But I wanna highlight some of the other advantages that you have in logical with, you know, doing your, you know, workflow. So I wanna highlight once again issue code. So issue coding is your friend. I I've had instances where a colleague says, we don't need to tag for these different requests for production. And, I roll my eyes because I know he's gonna ask me later, what are all the documents responsive to RFP for and expecting me to work a miracle to suddenly know that. It's like we're gonna have to code. Like, we have to tag if you want to know what RFP or interrogatory has responsive records because you can't just go, here's, you know, 4,000 email messages. You go figure it out requesting party. By doing that issue coding in advance, it empowers you to make production indexes. I've worked on a bunch of cases where we had to do production indexes, and it was a way to, push back on the requesting party when they said you didn't produce x. I could say, look at production index two that has everything that was responsive. Please read the production index that we gave you. Now there's debate with some of the how you read rule 34 in federal court. Documents need to be you know, have an index. Subpart is ESI technically doesn't if it's produced in the ordinary course of business, but there are many court orders where people are ordered to identify and create an index. And I would then also point back to someone in your office is gonna ask, what's everything responsive to RFP three? If you just tag that up front, it's gonna make life a lot easier in case this turns into a dispute later because it's gonna help you be able to, you know, command your data and know what's what. The other part of this is if you, you know, most likely are gonna retain something because it's privileged and because you're withholding it, you need to make a privilege log. That issue coding that's done, you know, you can say we withheld, you know, here these records that would have been responsive to RFP four, but they are attorney client communications and work product, and here they are on our production index or excuse me, our our privilege log. It could also be on the production index. But the point is you can make these, in logical. And after you you export out the CSV, you know, you do some work on it to make sure it it looks good. And, again, you turn the CSV into an Excel file, but it's a good way to, manage your productions so you can have a good production index and a good privileged log because I've, you know, had fights over those over over the course of my career. Now Ashu and I had slightly different experiences with this, and that just goes to show that not all practices are alike. No. I totally agree. But I will say one thing. Privilege logs are pretty universal in that sense, and there is a pretty easy format to produce them inside of logical, create a custom field for that privilege description, kick it out as CSV with those fields that you, you know, were obligated to produce to disclose why you withheld it, which is, I think, a great feature within Logical to say, hey. I can download a CSV with the proof description, whatever metadata fields I want to right here within perhaps two, three clicks of a button right here. Right? It's a great feature within Logical to download a CSV and very underutilized in my opinion. Yeah. I I will sing it from the rooftops. It is a great way to make your life easier because I did privilege logs at the beginning of my career just on an Excel file with a banker's box of paper before we were using any review application, and that was a traumatic experience. I don't want anyone to go through what I, endured as a young attorney. Now we have some fun questions coming in. And one is, when you're reviewing documents, is it hard to keep all the document requests, in your head as you review to make sure you were putting everything in the production index suggestions? The suggestion is issue coding, by being able to tag. So that way you're not trying to just rely on memory. Because I I would approach it from here's RFP two, ROG two, RFA two, if they are indeed related, and attack them as you're going through issue code. The challenge is you should review, you know, like, you've read all of your RFPs and written discovery before. So that way, you feel like, hey. This sounds familiar. And so because if it's responsive to one interrogatory, it might be responsive to another interrogatory because something might be responsive to five different requests. Well, you're not producing it five times, but you could be clicking the button five times to highlight the number of records it's relevant and responsive to. So that's one. And another is if you do a keyword search that identified a large set of documents, can ask help you clarify search terms to appropriately narrow results? Ash, I wanna pivot to you, but I wanna say yes, and you could. how we do that. Yeah. And I think it come down to the query or question you asked, but, really, you found too large of a set of documents. That would be the starting point of where I, you know, I said start broad. I would take then the keyword that got us those large set of documents. I'm only gonna interrogate those because that keyword hits on everything inside of that large set of documents. And based off what you're trying to find or craft, right, that additional instruction after answer provide 20 boolean search terms or provide, you know, boolean search terms related to x, y, and z. That's where logical excels, and you could copy and paste those kind of additional search terms right into that bulk keyword search builder that I was showing you and start seeing your hits or do your test searches and then build out a safe search from there too. Well said, because it's there are multiple approaches to this. Again, it's never one and done. There's no magic wand. It's you you will approach it. You refine it. You learn from it. You know, there were cases that I I've worked on where we we understood the breach of contract and an IP claim. And, you know, like, a month into doing document review, we've learned a bunch of other, elements of the case and, you know, types of old product numbers that were relevant. So that way the searches could be more refined. So, again, there there you learn context and, you know, you might double back because you've learned more, you know, when you're on, you know, RFP 15, because you know more than you did at RFP one and doing your document review. And that's that's normal. That's life. But with that, you know, again, there's a lot to this, that you can do to help you, practice efficiently and defensively in being able to say, I don't have to boil the ocean. I have 5,000 records. I've knocked it down to 2,000 to go through. And from there, I've been able to refine it and find what we actually need. And and with that, I think we're going to, you know, be wrapping up. Any closing questions for folks? This is the first time doing a product webinar like this, and so we definitely wanna get feedback and thoughts from folks, because this is new. And we do have some downloads for you under docs. We have a leveraging logical and responding to written discovery. We have a proportionally shrinking down, search terms and blog posts from yesterday that we just, download wrote. We have connecting the dots with written discovery, which is the companion webinar to this one. We also have the PowerPoint that you can download as well. And, we have with, with our creative team, a white paper on ask that, will be sent with a follow-up as well. So, again, this is a really exciting time. So any other closing questions before we turn it over to JQ? We'll let people ask, and, JQ, take it away. Awesome. Thank you so much, Josh, and thank you, Ash. This was an awesome conversation. It's I absolutely love using the UFO information. It's just so creative. It's a little different. I love it. It's been awesome. So it's really cool. I might have to dig through it myself a little bit later. Thank you everyone for joining us today. I appreciate it. As Josh explained, we've got a bunch of different documents. This webinar will be available on demand tomorrow. You should get an email if you attended here in person or even if you did not, you will still get a, an email that has that link to come back and watch it again. With that said, thank you so much. If there's any other questions, please fill out the survey if you have a second. And thank you so much for joining us. Have a great one.