WEBVTT NOTE Sentence-level transcript of https://www.youtube.com/watch?v=wzWNYDY7toc NOTE One cue per sentence. Cue ids are the line anchors on /transcripts/wzWNYDY7toc.html. A cue ends where the next begins, or 2 s after its last word. s1 00:00:00.000 --> 00:00:01.309 All s2 00:00:01.309 --> 00:00:03.309 [music] s3 00:00:13.519 --> 00:00:15.040 right. s4 00:00:15.040 --> 00:00:16.640 Hey guys, let's get going. s5 00:00:16.640 --> 00:00:18.080 Can everybody hear me? s6 00:00:18.080 --> 00:00:19.520 All good. s7 00:00:19.520 --> 00:00:20.320 Perfect. s8 00:00:20.320 --> 00:00:20.720 All right. s9 00:00:20.720 --> 00:00:21.760 My name is Jeff CS. s10 00:00:21.760 --> 00:00:25.199 I'm a staff customer engineer with DZ Labs. s11 00:00:25.199 --> 00:00:29.039 Um, as you can see from the slide, it says brought to you by Calibbra. s12 00:00:29.039 --> 00:00:32.160 Calibbra acquired DZ Labs last summer. s13 00:00:32.160 --> 00:00:37.990 Uh Calibbra is more of a data governance tool for structured and AI um agents and models. s14 00:00:37.990 --> 00:00:38.480 [snorts] s15 00:00:38.480 --> 00:00:44.879 So the gap that Calibbra had was we didn't have unstructured and that's really where uh DZ shines. s16 00:00:44.879 --> 00:00:47.920 So today uh with me is Leo Platzer. s17 00:00:47.920 --> 00:00:48.239 Yeah. s18 00:00:48.239 --> 00:00:48.960 Hi everyone. s19 00:00:48.960 --> 00:00:49.920 Uh I'm Leo. s20 00:00:49.920 --> 00:00:56.239 I was previously the CTO at DZ Labs which as Jeff has already mentioned was acquired last year by Calibbra. s21 00:00:56.239 --> 00:00:58.559 And you can see we're both twins as well. s22 00:00:58.559 --> 00:01:01.440 We have our hats that we're giving away at our booth as well. s23 00:01:01.440 --> 00:01:03.280 So come by and get them. s24 00:01:03.280 --> 00:01:04.960 First come, first serve. s25 00:01:04.960 --> 00:01:11.680 So today we're talking about um how we can go from a stalled PC to production with DZ data curation. s26 00:01:11.680 --> 00:01:15.119 So really focus on unstructured data. s27 00:01:15.119 --> 00:01:21.920 So our demo scenario I typically talk about is a manufacturing uh company, North River Manufacturing. s28 00:01:21.920 --> 00:01:25.200 You can see that you know there's 40 40,000 employees. s29 00:01:25.200 --> 00:01:27.520 They operate in 20 different countries. s30 00:01:27.520 --> 00:01:31.280 Uh really what they're trying to do is they have millions of documents in SharePoint. s31 00:01:31.280 --> 00:01:37.360 They've accumulated them over years and years and they need to figure out which ones are relevant for their chatbot. s32 00:01:37.360 --> 00:01:41.759 They're invested heavily in AI to improve operational efficiency. s33 00:01:41.759 --> 00:01:48.240 And their big objective this year is they want to build a legal ops AI chatbot. s34 00:01:48.240 --> 00:01:58.399 Now what they want to do is they want to help the legal and procurement teams answer questions around supplier contracts, partnership agreements, and it's going to be real time. s35 00:01:58.399 --> 00:02:04.320 So some of the questions they're trying to answer or what are the termination clauses for a specific supplier? s36 00:02:04.320 --> 00:02:14.319 Um which agreements expire this quarter, which agreements expire next quarter, and what partnerships have u exclusivity terms. s37 00:02:14.879 --> 00:02:19.120 So some of the folks in the teams that we've been working with um is Sarah. s38 00:02:19.120 --> 00:02:20.800 She's a head of data and AI. s39 00:02:20.800 --> 00:02:23.520 So she's responsible for the accuracy. s40 00:02:23.520 --> 00:02:25.200 Um also there's Jim. s41 00:02:25.200 --> 00:02:26.640 He's the AI engineer. s42 00:02:26.640 --> 00:02:29.280 He just needs to know which files do I need? s43 00:02:29.280 --> 00:02:33.360 I have tons and tons thousands and thousands of files in SharePoint. s44 00:02:33.360 --> 00:02:37.360 I need to find just the relevant ones with no sensitivity. s45 00:02:37.360 --> 00:02:39.760 Priya is the head of knowledge management. s46 00:02:39.760 --> 00:02:43.120 She wants to make sure that no sensitivity is leaked. s47 00:02:43.120 --> 00:02:51.200 um and make sure that if there if there is specific data that's uh classified that it gets redacted as well. s48 00:02:51.200 --> 00:02:52.080 Then there's Michael. s49 00:02:52.080 --> 00:02:53.760 He's a director of legal ops. s50 00:02:53.760 --> 00:02:57.280 He needs a chatbot to deliver accurate responses to his team. s51 00:02:57.280 --> 00:02:59.040 So he's from the business. s52 00:02:59.040 --> 00:03:01.200 And then finally there's Jackson. s53 00:03:01.200 --> 00:03:17.040 So Jackson's enterprise transformation uh lead and this is his flagship use case but also he has 15 other use cases that he needs to to deal with right so he wants to have a framework a repeatable s54 00:03:17.040 --> 00:03:34.319 successful process he doesn't just want to do it for one time he wants to do it for multiple different use cases so what they're able to do so far they were able to take 40 around 40 different files s55 00:03:34.319 --> 00:03:50.000 pre-curated files and build a small little PC small little mini pilot where they can do uh parsing chunking OCR embedding and then put it in a vector database and a knowledge graph and they got pretty good responses s56 00:03:50.000 --> 00:04:02.239 right seems pretty easy you can do that with 40 files but when they try to do it with many many different files with 80,000 plus files. s57 00:04:02.239 --> 00:04:06.080 Right now a sudden it's a little bit more challenging. s58 00:04:06.080 --> 00:04:11.760 And so some of the challenges they had were I can't discover the files that I need. s59 00:04:11.760 --> 00:04:16.959 I don't know where the sensitive data is in my on SharePoint. s60 00:04:16.959 --> 00:04:18.880 I can't trust the quality of the data. s61 00:04:18.880 --> 00:04:25.120 So my background um I used to work at both Informatica and data bricks. s62 00:04:25.120 --> 00:04:29.840 So for me data quality is typically the six different dimensions of data quality. s63 00:04:29.840 --> 00:04:32.960 Completeness, conformity, accuracy. s64 00:04:32.960 --> 00:04:35.440 But when you deal with documents, it's a little bit different. s65 00:04:35.440 --> 00:04:40.080 It's I want to see is there duplicate information, conflicting information, right? s66 00:04:40.080 --> 00:04:51.120 Leo's going to talk about he has a really good slide about showing why it's so important and also the freshness of the data and also everybody's talking about context, right? s67 00:04:51.120 --> 00:05:03.120 So what we want to do is when we tag the metadata, we want to use that in our chatbot in our vector database to make it more relevant, have better accuracy in the chatbot. s68 00:05:04.400 --> 00:05:08.960 So you can see some of the challenges that they had, right? s69 00:05:08.960 --> 00:05:11.440 They're getting incorrect answers. s70 00:05:11.440 --> 00:05:15.120 There's sensitive information that's being leaked out in the chat bots. s71 00:05:15.120 --> 00:05:18.880 It takes a ton of time to go to production because they can't find the files. s72 00:05:18.880 --> 00:05:25.039 is spending too much time looking for the files instead of writing code and doing things that they they really like to do, right? s73 00:05:25.039 --> 00:05:26.720 It's difficult to maintain. s74 00:05:26.720 --> 00:05:38.160 I'm working with another telecommunications company and they have six to seven different SME and they're trying to curate 20,000 HTML files and they can't keep up, right? s75 00:05:38.160 --> 00:05:40.080 So, it's very brittle, very difficult. s76 00:05:40.080 --> 00:05:45.120 And what we do is we provide the ability to automate this entire process. s77 00:05:46.240 --> 00:05:55.600 So over here now you can see they want to they want to try to access and leverage over 80,000 different files. s78 00:05:55.600 --> 00:05:59.120 So they went from 40 files to over 80,000 files. s79 00:05:59.120 --> 00:06:07.360 And really what they want to do is they want to find which files are relevant for their chatbot, which have high quality, which ones are up to date with knowledge. s80 00:06:07.360 --> 00:06:12.800 They don't want to use a a version where the contract is is old and expired, right? s81 00:06:12.800 --> 00:06:29.039 and they want to enrich their chatbot with the metadata that they've tagged and of course they don't want any sensitive data and so what we do over here is we provide the ability to do that right we can access the data s82 00:06:29.039 --> 00:06:42.880 um over here we can understand what the data is we can figure out which files are relevant do the OCR the chunking and things like that so let me show you a little a brief demo So s83 00:06:42.880 --> 00:06:54.560 over here you see in SharePoint much like what North River has they have a whole SharePoint site full of different you know there's call transcripts there's PDFs there's excel there's word documents s84 00:06:54.560 --> 00:07:07.360 and I they don't know which ones are relevant which ones are focused around legal now over here when we look at DZ we've already opened up the application right here and I can see you know I've already ingested 339 s85 00:07:07.360 --> 00:07:14.080 files and now I can see both the uh page page level or chunk level as well as a file level. s86 00:07:14.080 --> 00:07:18.000 I can see all the information right here. s87 00:07:18.560 --> 00:07:21.120 And then over here, this is what we call a texonomy. s88 00:07:21.120 --> 00:07:24.080 Taxonomy is a grouping of tags or a tagging tree. s89 00:07:24.080 --> 00:07:26.560 So I've ingested the 339 files. s90 00:07:26.560 --> 00:07:30.720 And now I can either manually create a userdefined tag. s91 00:07:30.720 --> 00:07:38.560 I can leverage tags that are reusable for my tag library or I can leverage AI to autosuggest what tags I should even have. s92 00:07:38.560 --> 00:07:38.880 Right? s93 00:07:38.880 --> 00:07:42.240 I can leverage AI to say you these are the tags are relevant. s94 00:07:42.240 --> 00:07:52.240 So here I just type in my tag name data document categorization give it a prompt and this prompt is what we're going to send to the LLM to determine s95 00:07:52.240 --> 00:08:03.440 is this file how should we tag this file now if I know what I'm looking for I can put in the classification values it's optional and again if I can leverage AI to autosuggest s96 00:08:03.440 --> 00:08:14.720 all these tags for me now also the other thing to point out if I have my uh my taxonomy here, my valid values um already. s97 00:08:14.720 --> 00:08:16.560 I can upload them through a CSV. s98 00:08:16.560 --> 00:08:22.560 And so once once we do this now, I can see I started to build this this tagging tree. s99 00:08:22.560 --> 00:08:26.319 Now the other thing I can do is I can have conditions on my tagging tree. s100 00:08:26.319 --> 00:08:29.039 So remember I said I only want legal and contracts. s101 00:08:29.039 --> 00:08:30.560 That's what they're looking for. s102 00:08:30.560 --> 00:08:35.120 Now what I can do is I can have have um subtagging. s103 00:08:35.120 --> 00:08:37.680 So only legal and contract files. s104 00:08:37.680 --> 00:08:41.279 I want to pull out more data around the contract type. s105 00:08:41.279 --> 00:08:42.880 Is it a supplier contract? s106 00:08:42.880 --> 00:08:44.720 Is it part of an NDA? s107 00:08:44.720 --> 00:08:45.200 Right? s108 00:08:45.200 --> 00:08:49.440 Is it materials and and time contract? s109 00:08:53.200 --> 00:09:01.040 And now what we can do when we add this, you can see we have different branches in our tagging tree. s110 00:09:01.519 --> 00:09:05.680 And I could have um unconditional tags as well. s111 00:09:05.680 --> 00:09:07.680 I can have any number of different tags. s112 00:09:07.680 --> 00:09:09.600 Now, what we're doing is we're going to generate the metadata. s113 00:09:09.600 --> 00:09:12.000 So, I choose what files I want. s114 00:09:12.000 --> 00:09:14.320 I can choose what tags I want. s115 00:09:14.320 --> 00:09:15.760 I can say generate metadata. s116 00:09:15.760 --> 00:09:16.880 And we'll run this. s117 00:09:16.880 --> 00:09:22.560 And you'll see the results over here. s118 00:09:22.560 --> 00:09:24.880 So now I can see the document category. s119 00:09:24.880 --> 00:09:26.160 I can hover my mouse. s120 00:09:26.160 --> 00:09:27.519 There's engineering documents. s121 00:09:27.519 --> 00:09:29.519 There's my contracts and legal, right? s122 00:09:29.519 --> 00:09:32.000 I have 68 different files. s123 00:09:32.000 --> 00:09:37.680 And if I click on any of those, it automatically filters for me. s124 00:09:37.680 --> 00:09:40.160 Now I can also see the evidence. s125 00:09:40.160 --> 00:09:41.839 Why did AI tag this? s126 00:09:41.839 --> 00:09:47.440 Where in the document did it find evidence that it's a legal contract? s127 00:09:47.440 --> 00:09:48.000 Right? s128 00:09:48.000 --> 00:09:49.920 And I can also do thumbs up and thumbs down. s129 00:09:49.920 --> 00:09:51.760 So there's fshot learning as well. s130 00:09:51.760 --> 00:09:54.959 So I can give it an example if I want to. s131 00:09:54.959 --> 00:09:57.839 Now you can notice I also have filters. s132 00:09:57.839 --> 00:09:58.880 So I have two different filters. s133 00:09:58.880 --> 00:10:04.160 I quickly went from 339 files down to 20 files. s134 00:10:04.160 --> 00:10:08.959 So, we also have the ability to scan for sensitivity as well. s135 00:10:08.959 --> 00:10:14.000 So, whether it's PII, PHI, um could be anything. s136 00:10:14.000 --> 00:10:24.880 I was in uh in London at a Gartner conference uh a couple months ago and there was some folks from Finland that came up and they wanted to know, you know, can we find Finnish national ID? s137 00:10:24.880 --> 00:10:26.240 I don't speak Finnish. s138 00:10:26.240 --> 00:10:27.200 So, I put it in there. s139 00:10:27.200 --> 00:10:38.640 I did this demo and not only can we do AI based, we can do patternbased with context and it ended up figuring out the the reax um syntax for me, but it also added s140 00:10:38.640 --> 00:10:43.680 some Finnish words which I thought were probably associated with Finnish national ID. s141 00:10:43.680 --> 00:10:48.399 So, it's going to be less false positives as well. s142 00:10:52.720 --> 00:10:55.040 All right, let me keep going here. s143 00:10:55.040 --> 00:10:58.320 So the other thing we can do is we can do quality. s144 00:10:58.320 --> 00:11:09.200 So again for us we have a dashboard in here for quality where we can go through and see what are the different groupings for um for duplication for conflicting information. s145 00:11:09.200 --> 00:11:15.440 Um and then we'll we'll leverage AI to auto suggest which ones to keep. s146 00:11:15.440 --> 00:11:23.760 Now the other thing we can do too is again for freshness we can say I only want files within x number of days or within a specific date range s147 00:11:23.760 --> 00:11:36.240 and then the last thing we can do for data quality is you know here's the the pieces of metadata that I absolutely have to have and then what we'll do is we'll figure out you know of the 339 s148 00:11:36.240 --> 00:11:47.200 files maybe you really need 200 files and then from there you can create a data slice and then we can also give it context as uh over here. s149 00:11:47.200 --> 00:11:54.240 So I'll go through one more video and we also have a live demo for you that Leo is going to be doing in a second. s150 00:11:54.240 --> 00:11:58.959 So in this case now what I can do is I can I can leverage a a data slice. s151 00:11:58.959 --> 00:12:04.079 So my data slice could be filtered based on the specific criteria that I want. s152 00:12:04.079 --> 00:12:06.560 So in this case I have 14 files. s153 00:12:06.560 --> 00:12:10.160 These are only the files that are relevant for legal. s154 00:12:10.160 --> 00:12:13.600 And now what I'm doing is I'm copying my use case description. s155 00:12:13.600 --> 00:12:23.279 What I can do is I can say here's what I'm trying to build and I can either do a flat uh tonomy or multiple layer levels in my tagging tree s156 00:12:23.279 --> 00:12:26.880 and then over there when I run it it's going to autosuggest these tags for me. s157 00:12:26.880 --> 00:12:31.920 There's human in the loop so I can choose which ones I like, which ones I want, which ones are relevant. s158 00:12:31.920 --> 00:12:36.800 But again, it's all done for me based on what I'm trying to do. s159 00:12:36.800 --> 00:12:42.079 We gave it context and based on the content of the files themselves. s160 00:12:42.240 --> 00:12:48.160 And then over here it even builds out the classification values for me. s161 00:12:50.800 --> 00:13:04.320 And then if we want when we run when we generate the metadata we're the job now we can see over here's all the sensitive information right here's all the different information that we have over here. s162 00:13:04.639 --> 00:13:08.399 And then again I can see the evidence at the file level and the chunk level. s163 00:13:08.399 --> 00:13:10.240 I can see the confidence score. s164 00:13:10.240 --> 00:13:17.360 So coming back from from AI what the model return I don't have to guess you know it's 100% confident 95% confidence. s165 00:13:17.360 --> 00:13:18.000 Here's why. s166 00:13:18.000 --> 00:13:27.519 So you can have very a lot of confidence in the the um AI ready data product that you guys are going to be building. s167 00:13:30.399 --> 00:13:32.480 All right. s168 00:13:33.440 --> 00:13:37.200 And now finally what we can do uh for delivery and maintenance. s169 00:13:37.200 --> 00:13:46.560 So you know one other thing that that's really important is with our data slice technology we have workflows where you can put on a set schedule so it's auto refreshed. s170 00:13:46.560 --> 00:13:49.440 So we want to make sure that your chatbot never gets stale. s171 00:13:49.440 --> 00:13:59.600 So if there's new records or sorry new files that get added to SharePoint or deleted, we'll automatically pick those up and add those on a set schedule to my s172 00:13:59.600 --> 00:14:10.959 um uh to my data slice or my AI ready data product and then our SDK will pick that up in the chatbot and then when we load the Victor database, you'll always have the most freshest s173 00:14:10.959 --> 00:14:13.120 uh metadata. s174 00:14:14.240 --> 00:14:16.320 So what's it mean for North River? s175 00:14:16.320 --> 00:14:21.680 So basically they were able to go from four months of data prep work to down a days. s176 00:14:21.680 --> 00:14:25.199 They were able to increase their data utilization for AI. s177 00:14:25.199 --> 00:14:33.920 They increase their chatbot accuracy because they not only were able to figure out which files to use but also the metadata from the files. s178 00:14:33.920 --> 00:14:39.120 They were able to inject that and leverage that through our SDK in their chatbot. s179 00:14:39.120 --> 00:14:39.440 Right? s180 00:14:39.440 --> 00:14:42.639 And then the other thing they did was they reduced their compliance risk. s181 00:14:42.639 --> 00:14:51.839 they knew where all the sensitive data is and they made sure that they were not part of they were all filtered out in their u data stores. s182 00:14:52.160 --> 00:15:00.160 Okay, so I'm going to turn it over to my brother from another mother, Leo, and he's going to talk about um context. s183 00:15:00.160 --> 00:15:00.959 Yeah. s184 00:15:00.959 --> 00:15:01.839 Hi everyone. s185 00:15:01.839 --> 00:15:11.440 Um I just wanted to basically like give this so what right technically okay you dduplicate your data, you reduce your search space, right? s186 00:15:11.440 --> 00:15:13.440 you clean it up, you have it fresh. s187 00:15:13.440 --> 00:15:14.959 What does it actually mean? s188 00:15:14.959 --> 00:15:15.279 Right? s189 00:15:15.279 --> 00:15:25.760 So, if we were looking at the vector space, right, when you ask a question, it comes in and you have duplicated data with the same information or potentially even s190 00:15:25.760 --> 00:15:29.760 a different version of the document with conflicting information now, right? s191 00:15:29.760 --> 00:15:31.360 Think of your Slack messages. s192 00:15:31.360 --> 00:15:34.720 Today, a person says this, tomorrow a person says that, right? s193 00:15:34.720 --> 00:15:37.440 The truth changes over time. s194 00:15:37.440 --> 00:15:49.920 If it ch if it stays however in your knowledge base in your search space your top K will always be full of pretty much irrelevant information right so in DZ for example s195 00:15:49.920 --> 00:16:03.759 something which is considered duplicative is either another piece of context like your database has changed and I can actually use it or it is simply redundant and we will remove it from the knowledge base. s196 00:16:04.320 --> 00:16:08.720 The second piece is like same as same reason as before. s197 00:16:08.720 --> 00:16:11.040 If your information is old, right? s198 00:16:11.040 --> 00:16:26.959 If you suddenly are working with data from before the 2000s, if you're working with information which was only relevant last week is not relevant this week, in the future some of the use case when agents actually get into production for let's call it the most high stakes use cases s199 00:16:26.959 --> 00:16:30.000 that has detrimental effects. s200 00:16:30.320 --> 00:16:44.720 And we have done um a lot of evaluation around what is the effect of making sure that your unstructured data quality is maintained, is consistent and is always handled s201 00:16:44.720 --> 00:16:48.560 deterministically in the same way, right? s202 00:16:48.560 --> 00:17:06.559 And what we're essentially seeing is that if your raw corpus, okay, let's imagine you have a thousand files and 30% of your files are stale or duplicative, it can happen that up to 80% of your actual context s203 00:17:06.559 --> 00:17:12.160 your agent uses to answer a question is filled up with stale information. s204 00:17:12.559 --> 00:17:23.039 which means like it actually uses 80% of completely redundant useless information and knowledge and this has to be managed. s205 00:17:23.039 --> 00:17:27.919 Um what we have seen is that by running on the exact same tasks. s206 00:17:27.919 --> 00:17:29.360 So this is a rack chatbot. s207 00:17:29.360 --> 00:17:35.919 It's not um running an actual like let's say cloud code task or codeex on a coding task. s208 00:17:35.919 --> 00:17:40.240 It was simply like multihop rag evaluation. s209 00:17:40.320 --> 00:17:50.640 And what we saw is that if we look at recall one like I mean we pretty much double uh the recall um and overall improve our accuracy. s210 00:17:50.640 --> 00:17:55.679 So like task completion by between 10 and 15%. s211 00:17:55.760 --> 00:17:57.440 And why is this so significant? s212 00:17:57.440 --> 00:18:02.320 This is so significant because this has nothing to do with the use case you're working with. s213 00:18:02.320 --> 00:18:08.080 This is true independent of what you're actually using the data for. s214 00:18:08.080 --> 00:18:08.640 Right? s215 00:18:08.640 --> 00:18:15.679 Because if you actually know what you want to use your data for, you can apply relevancy and all not other data quality dimensions, right? s216 00:18:15.679 --> 00:18:20.000 But duplication and freshness is true no matter what. s217 00:18:20.240 --> 00:18:27.679 And yesterday, right, there is a lot of context topic, a lot of context flags around here. s218 00:18:27.679 --> 00:18:36.880 Um, I wanted to give a very very quick demo on how you would use the DZ SDK to prepare a context repository. s219 00:18:36.880 --> 00:18:39.120 And I'm going to run this live. s220 00:18:39.120 --> 00:18:53.200 So what this is doing is essentially it's going to fetch all of the metadata we have about a given SharePoint site and we will then contextualize each folder within that SharePoint site as well as create an index s221 00:18:53.200 --> 00:19:04.160 over all folders so that your claw code for example or your codeex does not even have to look at the individual files to figure out what to look for but simply looks at the context s222 00:19:04.160 --> 00:19:06.320 MD file. s223 00:19:06.320 --> 00:19:23.600 So what you're seeing here is essentially really by using DZY you can curate context files where you are directly aware what exactly is within each folder right what is its purpose s224 00:19:23.600 --> 00:19:35.440 right when do you want to use it so for example this is one to one the mirror of what I actually have in SharePoint but now instead of actually looking at these 23 files s225 00:19:35.440 --> 00:19:38.160 figure out whether I even should look at this folder. s226 00:19:38.160 --> 00:19:47.520 I know exactly what document types I have, what it is about, what are the key topics, when to use this folder, and what are example questions. s227 00:19:48.320 --> 00:19:53.600 So this is how this is all the kinds of use cases today, right? s228 00:19:53.600 --> 00:20:03.840 How you can use the easy for use cases like your rack chatbot or use the easy to actually improve the performance and reduce, for example, the token cost for your agents. s229 00:20:03.840 --> 00:20:11.200 Um, but yeah, basically I'll move over to Jeff again to summarize this whole presentation. s230 00:20:12.000 --> 00:20:13.600 Thanks, Leo. s231 00:20:13.600 --> 00:20:17.600 All right, so guys, um, we're almost out of time. s232 00:20:17.600 --> 00:20:21.039 Our booth is way in the back by stage two. s233 00:20:21.039 --> 00:20:25.760 Please come by, bring any different use cases that you guys have. s234 00:20:25.760 --> 00:20:33.600 Um, we're happy to have discussions, do live demos, um, talk about what your challenges are around unstructured documents. s235 00:20:33.600 --> 00:20:34.799 um over here. s236 00:20:34.799 --> 00:20:40.080 Um before we leave, just want to let you guys know we have oneweek PC's. s237 00:20:40.080 --> 00:20:42.480 So, we're happy to prove our technology. s238 00:20:42.480 --> 00:20:45.919 Uh again, you know, you guys bring your your business problem. s239 00:20:45.919 --> 00:20:55.600 Um you're trying to solve with AI, bring some sample data, a bunch of different files, and then what we can do is we can get generate metadata for each different files. s240 00:20:55.600 --> 00:21:01.200 Um we could get a full sensitivity report on your on your data, data quality report as well. s241 00:21:01.200 --> 00:21:08.720 and then we'll show you how we can build a full tonomy based on your specific use cases, your specific files. s242 00:21:08.720 --> 00:21:13.039 So with that, everybody, appreciate you guys spending time with us. s243 00:21:13.039 --> 00:21:17.840 And lastly, we only have one more case of these hats, so come on by and grab them. s244 00:21:17.840 --> 00:21:18.960 So we'll be out in the back. s245 00:21:18.960 --> 00:21:19.600 Thanks, guys. s246 00:21:19.600 --> 00:21:22.000 Appreciate you.