Welcome to the the first webinar we're doing, on on ThoughtSpot from a from an EMEA perspective. And we've been out of this new ThoughtSpot with me. So the goal of this is just to give you introduction of the software, how it differs from other BI tools. We'll jump into the to to the software itself and show you some cool things you can do, inside. For those of you who haven't joined one of our webinars before and don't know who Interworks Interworks are, we are a technology consultancy, a data consultancy, that kinda specialize in consultant services all across the the data spectrum. So we have technology partners from data engineering, data data architecture, and then all the way up to kind of the front end data visualization. And ThoughtSpot is is one of our partners in that space. And I'm Matt. My email is there if any of you want to contact me after after the webinar. And I work work within analytics, so work within those those front end data visualization tools. So quick agenda for today. We're just gonna do a quick overview of of what ThoughtSpot is and how it differs from from other tools within similar spaces. And then we're gonna jump into the tool. We're gonna connect and prepare some data. We're gonna go over some analytics basics and and some stuff a little bit more, complicated. And we're gonna go through, what content is in Fort Spot, which is answers and live boards. And then we're gonna look at kind of some additional features that ThoughtSpot have brought in, recently as well. So first of all, what is what is ThoughtSpot? And for those of you who haven't used it before, ThoughtSpot is a search and AI driven analytics platform. And what I mean by that is it's purpose built for searching data and using natural languages natural language to query, and generate visualizations. And that differs from most kind of other visualization tools, which may, build visualizations through drag and drop or from code. This is all done by the user using that natural language query. And so similar how you would search on a on a site like Google, but you do it on your data to build your visualizations. And that means that Fortsbot empowers employees at all levels of organizations to make data difference driven decisions quickly. And that's because it takes away that need to go to an analyst or to a developer to get stuff, changed or added. It puts the power into your hands to create visualizations yourself, without having the technical technical knowledge that you might need when it comes to other tools. So to truly understand kind of what ThoughtSpot is, it's I think it's key to compare it to to other, other kind of dashboarding platforms or traditional tools that we might use for for dashboarding. Now when we think of dashboarding, we kind of think of prebuilt data content. It's it's a dashboard that's built for users to explore, maybe to add filters and parameters and actions, but it's not entirely customizable by the user. There's a limit to what the user can do. Our four spot is all about on demand data visualizations. It's about typing in a question, a query, and getting an answer instantly, that the user generates. And that's through something that is called natural language search. So that is referencing your data and your columns and your keywords in a search bar to build those visualizations. Whereas in traditional dashboard, in that process, is usually owned and maintained by developers or analysts, that, would would build those dashboards and workbooks and be the ones responsible for making any changes to them. And another key element of what we call search driven analytics is this concept of flexible drill down. Now in traditional tools, usually, if you were to drill down from your data from one dimension to another dimension, you might have to build that hierarchy manually, using groups, or custom actions. But, essentially, it's it's preset, and it's gonna be hard coded. Not in all not in all tools, but in in a lot of them. Whereas ThoughtSpot is all about a flexible drill down. You can drill down from one dimension into any other dimension that's in your data. What allows you to do that what allows you to do with that is that it doesn't mean you have to build it means you don't have to build a chart, for every different dimension that you want to see. You can drill from one into another, and really making that time to insight a lot quicker. And because the power is in the user's hands, you can all you can personalize your content. That means you can ask ask the questions that you want, and then you can share it with the people that you wanna share it with. A traditional dashboard might have you can answer some questions. You can set your own filters and parameters, but there's limited amounts of detail you can go into because you don't own that content to get on demand, answers to your questions in real time. To get on demand, answers to your questions in real time. Now that can be the case with traditional dashboard as well, but in the wild, what we see a lot of people is using kind of extracts of data, so generating a refresh of that data once a day or once a week to get a view in that time. Foresight is all about that live data, so all about seeing it in real time. Now if we think of a typical dashboard development process, here at Intuit, we usually think about it in around five stages. So So we look at the define, develop, refine, launch, and maintain. So when we talk about our first step, we define when we define a a a dashboard or a tool that we wanna build, we look at data requirements, and then we wireframe, and we maybe do some data discovery. And then we start developing. We build a version one of a dashboard. We look at validating some logic. We bring in maybe some design and UX and really kind of develop that data source to optimize it for for our dashboard. And then when we we just before we launch it, we'll probably look to get some feedback from from our key stakeholders based both on the kind of visual side and on the data accuracy side and make sure we're we're we're using that user feedback to to make it as good as possible. Then we launch, and we may add some training and some documentation and start sharing it around. And then it's this phase maintain where we maybe we action some change requests, we build some ad hoc reports, and we fix any bugs that have come up in in the production process. But what can often happen in this dashboard development process is you can get stuck between sections, and you can spend a lot of time in these loops. In develop and refine, we can get maybe too obsessed with validating logic and and making stuff pixel perfect, and maybe find it hard to prioritize exactly what to do. But the the thing we see most is this, this loop between develop and maintain. And this is because a lot of dashboards often have multiple diverse users. A lot of different groups in the business, a lot of different, user roles may have access to to to one dashboard, And they often have unique user requirements. And what that can lead to is that they can ask they could be asking for a lot of modifications. And often what happens is that these modifications to a user might seem very small. But as analysts and developers, we know that sometimes these small requests can actually take a lot of time, to build. And that kinda takes a time away from an analyst being developer being creative and building stuff that's useful to the business. You kinda get locked in this loop of, basically carrying out these these change requests for people that end up taking a long time to develop, but maybe maybe only getting used by one or two people. And then if not managed well, you you kind of then get a situation where a dashboard might get increasingly complicated. You might get increased training requirements. Objectives get lost in kind of a confusion when you're kind of trying to satisfy so many users, with one tool. And then you get into the the the phase of people creating workarounds. And and, eventually, this leads to dashboard being very hard to maintain and very hard to pass on to people if a specific developer leaves the organization or leaves the role. We kind of call this the endless loop of despair. This is constant maintenance period. There's a constant inflow of requests, and it's basically a never ending inbox. And we found the real problem with this, it takes it takes the analyst time away from doing useful stuff, from building tools that are actually gonna get used. They're they're spending time making these tiny little changes to satisfy unique groups of users, that might only kinda wanna ask one or two things on the dashboard. And that's where really ThoughtSpot comes in. Like I said, it really puts the power, of, generate insight in the user's hands. It it means that they don't have to spend a lot of time. It means that they're not gonna be bugging analysts constantly with with constant little modifications because they can ask these questions themselves within the platform. There's not a big technical gap like they would be if they were using a traditional tool. They might have to learn the logic, the language, the syntax. In full spot, it's simply asking questions. And then all this time then isn't spent, isn't spent maintaining, one dashboard. They can move on to move on to more useful things. So, yeah, we class for Foresbot as self-service search analytics. It's all about connecting to your enterprise data, searching using natural language, and then visually, visually, exploring stuff, embedding, and sharing insights across your organization. So, yeah, like I said before, search driven analytics. It's all about that search like experience. It's it's meant to be as comfortable for a nontechnical user as possible, to get the data that they wanna see and visualize. And then once you get that data, we wanna drill into it. We wanna customize it, and we really wanna share it across the organization. And the ideal situation is then that leads to eliminating most ad hoc SQL or ad hoc custom code that gets written to satisfy small, group of people. And something that, I'm sure that anyone in the data space knows that AI is a big word right now, and it's gonna continue to be a big word and a big feature in the future. Last year, ThoughtSpot introduced its its AI and natural language capabilities. So this allows you to to basically search data using, kind of real world English rather than data speak. And I'll get I'll get onto that later when we demo the product. But, essentially, it it it makes that technical gap even even less, essentially, because because it allows users they don't even have to reference the exact columns or keywords. They can they can really search as if they were searching, Google. And then like I said before, yeah, this is a drill into data concept. It's all about drilling down into any dimension you want. So, again, putting the putting the power in the user's hands to to to to let them get the insight that they want rather than trying to spoon feed it to them. And because the power is in the user's hands, they get to personalize what they see. They can have personalized KPIs that you can share with maybe just your team, a wider group, or just keep them to yourself. And you can put those cut KPIs front and center of your home screen to always look look at when you sign into your Forcepoint instance. And then once you've got your your visualizations and your charts, Foursquot is all about sharing. It's all about keeping take taking out data outside of Foursquare as well. It recognizes that in the modern world, you need to connect to a lot of different tools. So it allows you to share across applications like Slack and Teams, Salesforce, and Google Sheets. It doesn't make you keep your data inside full, but it encourages you to share it in other platforms as well, which which is very key. So that's a quick overview of the platform itself. But I think to show it properly, we're gonna have to jump in. So I'm just gonna share my other screen in a sec. So this, for your views those of you who haven't seen it before is the the ThoughtSpot layout. So it's all in the browser, and this is what we call the home screen in ThoughtSpot. Now you don't have to download any desktop application or run it behind any firewalls. It's all available, to sign into your instance in in the browser of your choice. Now when you sign in for the first time, you'll usually go through a little bit of onboarding process. It involves looking at video and and showing you exactly what to do. Once you get past that, you can get to your home screen here. And your home screen is essentially an overview of all your content, and it's very much designed in a way that is meant to be friendly for users of all technical backgrounds. It's meant the idea of it is to look different from your traditional kinda server environments, but maybe look a bit more like a developer application, rather than a human application. So as you can see, we have kind of this this this social media type feel. So we have trending answers and live polls on the right hand side. This is kind of the most popular content in your instance right now, so you can see what your users are actually looking at. And then you can see kind of a full list that is searchable of all your content, inside Flotspot. And here you can also add KPIs to your watch list. So like I said, with personalized KPIs, you can add them to your home screen here so you can see them as soon as you sign in. And these aren't static views. These are live, live KPIs that will change as your data changes. So all of this allows you that customizable feel that you don't necessarily get with get with other tools. Now if you have a four spot instance, feel free to follow along, with this. Hopefully, when you sign in, you'll see something similar along the top. Depending on your permission levels, you might not see things like develop, and SpotIQ, but you should at least have answers, life boards, and search data. So like I said, the main feature of ThoughtSpot is its natural language querying. And if you are in ThoughtSpot, you should see this this big gray button here, and that's where you can search on your data. That's where you can build your visualizations, using that natural language query. So if we click that, we'll get taken into a separate screen, and this is our this is our interface for searching data. Now by default, Faultsbot will remember the last dataset you searched on, and it will select that when you go into your search data. But you can easily change that just by clicking down and selecting any data source that you have access to there. So a quick example of how it works. Like kind of a traditional search engine, you have this suggested search feature. So it will automatically as soon as you click on this little bar here, which is our search bar, you'll have fields that have been used the most recently, so your popular fields. And when I when you start typing, those fields will show you relevant ones to the the the words and letters that you're typing in. So in this case, I'm just gonna type in sales, and I'm gonna type in region. I'm gonna click go, and I'm gonna get this view that's instantly built out for me from those two words that are referenced. So that's a very quick demo of what to do. But first of all, I'm now gonna go back and show you kind of from the beginning how do you actually get data into FortSpark, and then we'll we'll jump into this space a little bit more, once we get a better idea of of of the data side. So to do that, you can go into this data tab here. Now depending on your levels of access and what an administrator grants, you might not see this. You might not have have it available. But, essentially, this is where you can see all the data sources available to you, within your Fortisport instance. So it's pretty much like a data home page. Now from here, you can create a new connection. So if I was to click this blue button, you don't need to follow along in this bit because I'll just be connected to our internal instance. So when you click that create a connection, you'll see all the inbuilt data connectors that Forespot has. So you have your, your typical cloud data warehouses, your Snowflakes, your Google BigQueries, your Databricks, and then your traditional ones as well, like Postgres and SQL. So all of these are all the the native connectors, that ThoughtSpot has. So from here, you can select one, and you can select your name. I'm just gonna call this test demo for connection. You can call it anything and give it a description if you want. And I'm just gonna click continue. Now from here, you can sign in to your instance. I'm just gonna sign in to my, Snowflake instance very quickly, just to show you what that process is. And when you click continue, you should then get a list of tables from your database. And from here, you can see any of the ones you have access to. So in this case, I'm just gonna go down to my solution sandbox. I'm gonna go to Superstore. For those of you familiar with Tableau, you might recognize the Superstore dataset. I'm just gonna click Superstore here. What I can now have is, all the columns from the Superstore dataset. I'm just gonna click here, and I'm gonna add those in. And I'm also gonna click Superstore people. I'm just gonna select these as well. So I have two tables from this connection. And when I click create, It should be fairly quick, but now I should have my tables, come into. Sorry, Matt. There's a question in the chat. Sure. David David, he asked, does such data need special permissions? So, yeah, there's a permission group for for for each kind of similar to other tools, there'll be permission areas for, each each thing. I don't know if there's specific one for search data, but in general, you would kinda need I think it's for view data permission, to be able to search on on stuff. So you can limit it if you don't want people to do that, but, generally, you want people in the platform to search data. But, yeah, you can limit, permissions, very granular like you can in can in most tools. Great. Yep. Gary, do we have labs? Don't think we have, lab session this webinar. No lab session this webinar. No. But it's something that we can, okay, keep in mind for for future sessions. Though when I do the when I do a demo inside, search data, if you do have an instance of Forcebot, you can follow along because I'll just be using the, the demo dataset. So once you, once you've brought your tables in, you'll be able to see them here. So you've got just Superstore and Superstore people. And what I can do now is select my individual tables. I can go to joins here. So first of all, I can see what all these columns are. I can see a description, data type, etcetera, and I can make any changes I want here as well. And I can also go to joins. At the moment, I have no joins yet. So if you have a database that has kind of a foreign key and primary key setup and you have tables joined inside your database, they will automatically pull through. In this case, I don't. So I'm just gonna add a join myself, and I'm gonna select my Superstore people. And I'm gonna select my region. I'm just gonna select my cardinality as many to one because there's many regions on the left and just one region on the right. I'm just gonna create my join. Once you do that, you can see this nice visual interface. You can see the Superstore is joined to Superstore people by that region key field, just to give you a visual dynamic. And you can have as many joins as you want. You will just add them from this from this, section here. So now I can use Superstore and Superstore people because they're joined together. Now one thing that's really useful in Foursquot is this concept of a kind of a semantic layer. So you have your tables directly from database, but what you might wanna do is optimize these tables for search. And what that means is potentially, you might have tables where the the names are a bit off, or there might be a lot of ID fields. They might feel they might be fields that aren't very, very friendly essentially to an end nontechnical user. You want fields that are kind of in a in a good condition to be searched on. So what Forcebot allows you to do is essentially create a semantic layer. So what I'm gonna do, I'm just gonna click on create new. I'm gonna go to worksheet. And from here, I can choose my my sources to include. Now I'm just gonna click Superstore and Superstore people. There are a couple that I've brought in in the past and click close. And now I have my two tables here, and all the columns available to bring in. So I'm just gonna click the top of Superstore and add my columns. And in this case, I probably don't need customer ID or product ID because I have the names already. And this can be a little bit confusing potentially for end users who don't know a data source that well and doesn't really bring additional value. So I'm just gonna do is delete these. And now I can save my, save my data source. I can just call this test. And now this is avail now this has been slightly modified. It's available to be searched on. And what we can do here is also make any additional changes to to columns that we want. So if we wanna change any measures to attributes, we can here. We can't change attributes to measures. That'll be set at the measures. That'll be set at the the data source level. But if you want a measure to actually be an attribute, you can here. You can change default aggregation, and you can change geo config as well. So a state, for example, I know that's a geographical field. So I'm just gonna click on that geo config, and I'm gonna change it to state United States. So you can make any additions here. There's a couple of other fields as well, but we we won't get to today's little bit more more more more in-depth, but there's a number of things you could do to these to these, to these, fields. So I'm just gonna save changes, and now this worksheet is available for for anyone with access to search on, within within this instance. So that's essentially how you connect to and prepare data. It's it's fairly simple. And then if you go back to the data tab, you can see all the list of your worksheets and your tables as well. So, yeah, fairly simple to bring data in and do any do any changes you want on it. Now that we we've done that, we can go back to the search data button. And I'm just for those of you who do have an instance and wanna follow along, I'll just I won't use the Superstore store dataset we bought. I'll just use the sample retail apparel, but it's very, very similar, because that that is a default, test dataset involved in in instances. So I'm gonna go back to that, and I'm gonna start doing some some searches. So the last time we looked at sales by region. I'm just gonna do that again to show you how it's built. So FullTrail will automatically build the chart that it thinks is most appropriate for the query you put in. So in this case, sales by region, it thinks that the column chart is the most appropriate for that. So if you do one measure by one dimension, it's usually gonna pick either a, a column chart or a or a bar chart. But that isn't set in stone. There's these options on the right hand side here. And if you select here, you can change your chart type to anything that isn't grayed out. So anything that is grayed out, if you hover over it, it'll tell you why. So a line stacked column, for example, would need, one attribute and two measures. So if you wanted to change it to a bar, you could click here, and, yeah, anything appropriate can be changed. You also have this edit chart configuration, and this is where you can change any elements of a view that you've created. So in this case, if I was to go to the y axis and click on total sales, I can change color. I can add some data labels. I can change the number to a currency. And, yeah, anything that you'd expect to do, you can do it in in this in this part here. And something to note is that ThoughtSpot has kind of a built in default color palette, and you might have noticed that it will change the color each time you bring, each time you create a different view. It kind of randomly selects a color. You can change this if you want, in your administration settings, if you just want one default color to start from, or you can also bring in your own color palettes as well. But I think it's quite a nice feature that is you're not you're not stuck with the same colors every time. So that's how you kind of configure the chart and change any colors that you want. Now I'm just gonna show another view quickly. So I'm gonna go to sales, and I'm gonna go to date. And as you can see, the date is date fields are always pink, attributes are always blue, and measures are always green to split them. And you might have noticed from the from the left hand side, there's three options to view your columns. So you've got this one in the middle, which I leave as my default, which is all of the fields split by measures and attributes. But if you have, like, a very big dataset with a lot of columns, you might wanna click on this one, which shows you the most popular columns used on this dataset. So if there are a select few columns that you use all the time, you don't wanna always look at a full list, you can use that option there. You can also use an a to zed if you like to see everything in the list, alphabetically. So you see we've got our sales by monthly date. I'm just gonna change it to a line chart quickly. And you can see the date I brought in is the has defaulted to a monthly date. And, again, this is something ThoughtSpot will infer from the values in your data, but to default aggregation essentially for the date, and it will build it out accordingly. So I've got sales by month. And something that's I find really powerful within Foursport is is the KPI cards. Now a lot of softwares, to create kind of nice looking KPIs of all the information that you want can sometimes be a bit challenging. It can involve maybe merging two charts to look like one, building kind of maybe a little complex percentage versus previous period calculation. The in full spot, you can actually just select the KPI card from this chart type. So if I click on my change visualization type here and go to KPI, it automatically builds me this, total sales by by monthly date KPI card. So it shows my my, my sales over time by month, but it also shows me kind of what my latest, sales are for a month and then compares them to a previous month without any kind of logic needed to be created by the user. It's simply just type type it in my sales by date in the query and then change it to a KPI chart. And from here, you can also customize it. So you can click on the configuration, go to settings, and you can change it essentially to to whatever you to, the built in option. So I might wanna actually compare it to the previous year, say month, and I might wanna change what is a positive or negative color. I might wanna not show anomaly bands or anomalies. You can do all that from a setting. So no need for any logic or any, any kind of merging of two sheets. It's all done kind of just by one search and, you know, slightly edited and changing the chart type to a to a KPI. Now first, you haven't used it before, you might be wondering, so how do I how do I filter data? So every visualization platform will have a way to filter filter data. Again, in Fortsbot, it works a little bit differently. So in in other tools, you might need to if it's code based, you might need to write a where clause or, select from a long list of data. Again, in Fortsbox, it's all in all in the search bar. So all I really have to do, I can I just have to reference my values that are in a column to search to search by? So let's say I wanna filter by region, and I know I only wanna filter by the east region. I can just type in east. Click go. Now that's the filter is categorized by this gray color. And just by typing in east, this is filtered by that east region. And I could go even further. I could type in the state I want. So New New Jersey, for example. And I've got a a chart, basically got the KPI chart over time, monthly, filtered by Eastern New Jersey just by typing in typing in forwards. So a very quick way to generate a top chart with two filters. And for those of you who do don't don't wanna do that or might not know the values of, the data you wanna type in, you can easily do the traditional method as well. So you can select any field on the left hand side and select the filter button. And you can also once you've typed in once, you can also edit it from here. They'll all the filters will always be above, the chart. You can go in here and select it the same the same. So that's a quick introduction to to filters, and, yeah, fairly flexible because you can still leverage the the natural language querying with with filters. So I'm just gonna take this off a second. Just go back to the original view. Now so far, what we've done is just reference our columns on the left hand side. So all everything we've typed in has been a reference to, a column name, essentially. What we can do in Fortsville where you can take your searches to the next level is the concept of keywords. Now what keywords are, words aren't in your data, but they're keywords that help you define a search. So these can be things that in other tools you might build build into the calculated fields or formulas. So things like top x, top n, if you wanna rank things. You have a lot of data keywords, like last year, this year, next year, and a lot of string words as well, like begins with tames, etcetera. And this this is where you can kinda take your searches to the next level. And if you, wanna see a full list, they are just on this this Fourspot documentation here. So there are a lot, and you don't need to remember them all, but they are all listed here. All the keywords you can type in to help you help you define your search. So I'm just gonna go through a couple of examples of those, and show how they can be how they can be powerful. So here I've got my sales by date, And what I might wanna do is I I I don't really care about maybe twenty twenty two. I wanna just look at maybe the last twelve or thirteen months. So I can simply go here and leverage this last keyword. So last isn't in my data source, but it's a keyword that's recognized and built into Foursquot. So what I'm gonna say is last twelve months, and you can see it recognizes that as a keyword. And now my data is filtered from the the fourth the first of, April twenty twenty three to first of April twenty twenty four. By doing that, it does it only it's only gonna show the completed months. But, yeah, I don't have to kind of do any date diff logic or any calculate fields. It's all there in in in this keywords. And but it doesn't have to be twelve. It could be I could edit this to be any months I want. So five months, and it works, works exactly the same way. Honestly, this isn't is now unavailable, the data point, because I've in my settings, I was comparing to the to previous year that now doesn't exist within five months. But, yeah, date keyword is very powerful, and there's loads of them. You could reference a specific day, date part, anything, to to to search in. So a couple of other keywords I find useful are the concept of using kind of top and bottom. So I'm just gonna do I'm just gonna do product by sales. And you can see here, this is my sales by product, and it's not a useful view because there's three hundred and forty five data points. I'm not really gonna get any insight out of that because I I'd have to scroll along, and and to kinda see who it was performing well and performing poorly and badly. So what I can do is actually this concept of top and bottom keywords. Now if I just type in the word top, it actually recognizes it without without having to type in a number. So by it's built in to show the top ten if we just type in tops. If I click that, it now automatically shows me that the top ten products ranked by sales, just from typing in that word. And if I was to change that to bottom, it would show me a bottom ten. And, again, by default, it's ten, but it doesn't have to be. I could do any random number, and it can show me the bottom thirty two products, for example. So it eliminates the need to kinda create maybe a rank calculation and and and scroll down a huge list or add a add a parameter to show to show bottom x at all. Could all be done within this within this natural language search bar. So I'm just gonna go back to the line. Sorry. We have another question from Oh. A, and he asked, would the nontechnical users have to type that field out to make sure they're looking at only complete sales? How would you make it easy for the nontechnical user to explore the data and make their own views? Yeah. Good question. What I would usually do in that case is is have that flag somewhere in the database, so an incomplete month flag or something. And then you can build that into either your worksheets and add as a filter, or you can put it put it as a field in here and add it as add it as a filter within your search data bar. So you have an incomplete month flag and only show kind of true and false. But, yeah, probably best to kind of put that in your in your database, so that that that the user doesn't actually have to do anything in here. You can just add it as a filter if needed. So I could say product by sales and then complete moves, and it would only show that if that if that existed. That's probably the the best way around that. Yeah. Add that flag in, from the back end. So, yeah, if I just look at my my product by sales again, we've got our our huge view. And there's also as mentioned, you've got your your number keywords, your date keywords. You also have your your string keywords as well. So I could say something like product ends with let's just say hat. And now I can see a full list of just my products that end with the word hat. So not messes not not massively useful for this example, but, again, something that you can you don't have to kind of go through a full list of filters and just search for word hat. You can just do it within the search bar. So these are kind of an example of a of a date keyword, kind of a top x keyword, and then there's a there's something you could do with a string keyword as well. Doesn't have to be ends with. It can be begins with similar to, all works for the same way. I'm just gonna clear this view again a sec. I'm just gonna show my sales again, and this time, I'm gonna say by item type. So another keyword that can be quite useful is the concept of comparing one value to another. And I have this few here. It's just sales by item type. Maybe I wanna see, in some cases, how some states are performing against other states. Now I'd and I don't really wanna bring in my state view in and then filter all the ones I want. I can simply do this by writing, there's a word called the versus keyword. So if I was going to write in let's say I wanted to compare New Hampshire, and I'd say versus Idaho. Click go, and then this will automatically build me out a chart for just those two states against each other colored, and then by that item type. And I can even do more. So I could do versus, let's say, California. So it's not limited to two. And instead of, again, going down a long list and filtering out, I don't like this view, so I'm just gonna change it back to a bar chart. And as you can see, you've got to cover these three axes that look look a bit messy. So what I'm gonna do here is I'm just gonna group them, which is similar to kind of synchronizing the axis. So I'm gonna group them to New Hampshire there, and I'm gonna group that one to New Hampshire. Sorry. That one needs to be California. Yep. And now we've kinda got one access for all of them fit on, so you can compare them three. So, again, no need to go down a long list of filters. You can do it all with just typing in that versus keyword and referencing your value. Now those are all kind of simple examples, but you can kinda take keywords to the next level by using some that are a little bit more complicated. So sometimes you might have a view like sales by date, shows the the the sales over time. Sometimes you might wanna see how those sales are performing month on month, how how are they growing, and how are they, how are they, if they're falling. So what we can actually do here is leverage a keyword called growth of. So we don't need to create kind of a percentage difference calculation. We can just do growth of sales. So I'm gonna say growth of sales by date. I'm gonna say monthly. So I'm I'm referencing my sales, my date, and I'm putting in a keyword of growth of and then monthly at the end of date. So it's just one. I'm not I don't even actually have to bring in any of the fields. I just have to click that. And let's just change it to a colon quickly. And now we can see how they're kind of going up and down by month. And you see this this is quite an outlier here from the from the day our our month started, essentially. So I think the first month of data is December twenty twenty one, so it's it's not a full month. So January is is very much much bigger than the rest. So what I might wanna do here is just say date after January twenty twenty two. And again, it recognizes a date and after as a keyword, and it recognizes the month and the and the year. And if I click that, I've got a much kind of nicer looking chart now. And, again, if you were to have that incomplete month flag, you could also break that in, and that would eliminate April as well. If you wanted to do it manually, you could also click on the date. Once I took if instead of that, and you could say date between January twenty twenty two and April and March twenty twenty four. And that would limit that incomplete month as well. So that's kind of some of the keywords and how they can be powerful. Like I said, there is so many more, and it's worth going through if you're if you're gonna use it just to to see which ones might be useful for yourself. Right. So that's Sorry. Eric has a question. What's the difference between a keyboard and token, and how slash if are they created differently when we build the semantic layer. So keywords are just unique to the search search data part. So they're just something that helps with this natural language, the the natural language part of the search. They cut you can't reference the keywords in in in that semantic layer. They are very much unique to the end user, essentially. So they're just there for that flat front end search capability. So we've created a few views. And we've we've referenced a few keywords. But one of the key kind of concepts of of ThoughtSpot again is about also all about about sharing sharing of data and and getting it out to the users that you wanna see it. So, again, we're just gonna go to sales, and we're gonna go let's go product again. Let's just use our top keyword again. Just click go. Now once you've created something you you like and you wanna share with other people in your organization or save for yourself for later use, we can save it as an answer. Now what an answer is in Fortsbot is a standalone visualization. So it's just one search, that can be saved as an answer for you to go back to. So for those of you who might come from a tablet background, this is equivalent of your worksheet within Tableau where you just have one visualization, one standalone visualization. So once you create that view, if you click on these three dots here, you have the option to save. When you save it, you can choose a name. I'll just leave it as total sales of product now. You can put a description as well, and you have the option to make it discoverable, which means that users who are part of the same user group as you and have access to the data source will get access to that answer automatically. If you turn that off, it means they don't have access, and you can just share it with with individual people as you like. I usually no. Most of the time, you wanna make something discoverable for people in the same group as you, but if you don't, you have option to turn it off. From there, you can save your answer. And what you can do is you can go to your answers tab along the top, and you can see the the the answer that you saved will be here. And you'll be able to also see all the other answers that have been created that you have access to. And from here, you can click it. You can share it if you wanna share other groups than yourself. You can apply a tag to it. And what a tag is in ThoughtSpot is essentially a way to, catalog content. So I'm these tags have been created by our administrator, and I could add any of these tags so people can when they search for stuff, they can search in the tags to limit, the amount of things they're searching for. So I could type this with Salesforce, for example, and then that I can now search that tag in my home screen to only look at SFDC tags. So that is an answer. So that's one part of the content within ThoughtSpot, and the other part is live boards. Now live boards are more similar to your traditional dashboards. These are a collection of answers, on on a page where they interact to each other. So, yeah, similar to your traditional dashboards. And what you can do there is that next to those three dots, you also have this big pin button. What you can do is you can select pin, and you can add it to any any live ball that your user has access to. In this case, we can you can also create a new one, which we're gonna do here. So I'm just gonna call this, demo webinar liveboard. And I'm gonna pin it to, my liveboard. And that's the equivalent of dragging a worksheet into a dashboard. It's just another way of it's it's just Foursports' way of saying it, essentially. And now if I go to my Liveboards, again, similar to answers, you'll see a full list of content of everything you have access to where you can also share and tag from this from this screen as well. I'm just gonna go into one I've just created, this demo webinar live board. And here I can see this this total sales by product that we've just created. So this now sits in this kind of, interface with space for other stuff to get, put in as well. So I'm gonna do I'm just gonna go back to search data. Let's just build another quick view, and let's say sales by store. Top twenty. Let's just add some data labels. I'll keep it as purple. I'm gonna pin to my, light bulb as well. By default, the last one will will stay. The last one you've used will be the one you can automatically pin to, to make it a bit quicker. So now I've got kind of two two views. I've got my my sales by product and my sales by store. Again, yeah, I can add as many views as I want in here. I don't really I don't think there's a limit. And what's what's good about it is that if you have one with a scrolls all the way down, they're not gonna kinda render until you scroll all the way down. So they kind of render on on demand, essentially. And to kind of once you've bought stuff into your LiveBoard and you're happy with it, you can then go in and kind of edit features of of the LiveBoard. So you've got this big edit button up here. I'm just gonna click that. And from here, we have our we have our options to to change stuff around in our LiveBoard. So you can resize anything just by using using this here. You can also change the default look. So if you want it a bit longer, you can do that here as well. But you can also use that that dragging and dropping interface to kind of resize it here. We also have our traditional options that you'd expect in a in a dashboard as well. So we can add a filter. In this case, let's just filter by region. And we have our default behavior that we want. I'm gonna leave it as multi select for this purpose. We We can choose where it's applicable to. So if we only wanted it to to choose by if we only wanted to filter by product and not store, we could untick that there. We also have a concept of link in filters. So if you had another data source that wasn't joined to your data source, that had fields in common, like a date, we can use those to link those two filters. So, again, for those of you from a Tableau background, this would be very similar to blending and using one field, across the two data sources that aren't actually, linked up at a data source level. So for now, I'm just gonna click apply and leave leave region there, to be selected. I'm just gonna leave all of them by default. So we can also add tabs. So like like a traditional dashboard tool, this doesn't all have to be in one tab. You can have different tabs for different where you're showing different types of data. That's all possible using that, this tab button here. And you can you you can add notes as well. So if you wanna put in kind of images that are, you have in your computer or text boxes, you can do all that from from here. If you have parameters that you've, that you've added as well, you can also add those the same way as you'd add filters. We don't need to get to parameters. We won't get to parameters state, but they essentially work very similar as they do in in other tools like Power BI and Tableau. It's simply changing, a value using, the date string or or or integer word, selection. So I'm just gonna save that here. We've got a very simple live we've now got a filter that applies to to both. So if I was just to select south and southwest, the two charts will will change accordingly. So I'll change it back. Now in terms of other features you might expect on on traditional dashboards, you can also filter using, the values in the charts. So if I was to right click on Colorado, I could filter here, and my other chart would filter as well. And I can simply right click to remove the filter and to go back, to how it was before. So those are some of the kind of traditional features that you'd expect. I'm now gonna go through some of the kind of things that are unique to FortSpark, and and where you can kinda get value added. And, again, I'm just gonna right click on a mark here, and I can do this on any chart. So any chart that has a mark has the available has the option to right click and drill down. Like I said before, when I was going through slides, you can drill down into any dimension. So I haven't set up any custom hierarchies. I haven't set up any groups, but I can click on something like Colorado, and I can drill down into anything I like. So we can see that Colorado has the most sales by state. Maybe I wanna see what makes up those what's making up those sales. So I'm just gonna drill into item type, and I can see that, okay, jackets are doing well. Kinda makes sense. It's probably a cold state in the in the winter months, so it's gonna sell a lot of jackets. So from there, maybe I wanna look drill down into jackets specifically and see what's selling well. So I can drill down further into product. And now I can see that okay. With from jackets in Colorado, it's like use M and T jacket that's really driving sales. I can and then to go back, I can simply undo. And maybe I can drill down into into something else. I can drill down into any column. Time. And like you can see, it's in those winter months that they're really getting sold. And like you can see, it's in those winter months that they're really getting sold, which makes sense. You know, it's gonna be very cold in the winter. So we don't have to build those extra charts, and we don't have to spend any time bringing them in and build any custom actions. They're all just available, to to to drill down into, and then you can easily just just go back to your original view. And you can do that on any any chart that has that has a mark. What we can also do is a bit less, bit less sexy, but we can also show underlying data for any data point as well. So we don't have to waste the time to export and stuff. It's all there ready to go out of out of ThoughtSpot if you wanna see, this in a table form and share it with people. It's not trying to keep data in in Fortsport. It's trying to get you to share in any way possible. So it's very quick to kind of show our underlying data and, be able for you then to to export it into a different platform. We also have some other cool features. So, like, if you want to a lot of the time, we we use dashboards, and we we get them out of our systems, and we put them into, presentation format and adjust all the sizes. Forcepoint actually allows you to present as a PowerPoint as well. So you can you can simply just click present, and it's there, to present as if it was a, a a a standard presentation, which can be quite useful, if you don't wanna kind of take a data out of the system and get into a specific format. And you might have noticed up here, we have this, Tappos SpotIQ. And I'm just gonna go over quickly what that is as well. So you can click again. You can click any mark, and you can click SpotIQ analyze, or you can also click the whole chart itself and click the SpotIQ analyze. And what SpotIQ does, it it shows you exactly kind of what trends and what things are making up this number or these numbers in a specific chart. So it looks at all your other dimensions and automatically comes up with an insight as to why those numbers are high, they're low, they're showing certain trends, and it does all this automatically in the background. So that was to click total sales by store and go to SpotIQ analyze. It asked me what I wanna know about my total sales by store. So it'll ask questions for me and help and automatically analyze my data. So I can choose whether I wanna see outliers, I wanna see trends, or I wanna see any cross correlation. I can click continue, and then it gives me those columns that it wants to analyze by, essentially. So those are the things that it automatically thinks that they're gonna be useful to show what makes up that number. If you wanna add more, you can, but things like zip queue zip code and latitude are probably not gonna be useful for this this type of analysis. So once you're happy with your selections, you can click analyze. And what happens is in the background, SpotIQ will get to work, essentially. So once it does, you can go to, your SpotIQ tab. And if I refresh it, there should be another one running in the background. Oh, no. It's already done. So it's it's analysis for total sales by store. Said it's a very small dataset, so it takes ten seconds or so. A longer dataset might take a couple of minutes, if you're running for for kind of millions of rows. Once you click that, you can go into kind of the the the analysis itself, and it will show you kind of why it thinks certain results are happening. So it knows that, for Nevada, there are certain products that are doing much better than others, which make up that number. For Indiana, again, the same thing. So it goes down all your possible dimensions to try and give you more insight as to why why those sales numbers are of a certain value. So but, yeah, instant analysis, nothing you have to do, all done in the background by by Force by itself. So that's kind of a a whistle stop tour of of LiveBalls. One more thing I'm just gonna show quickly is that you might notice this explore button. And what this does for user, it allows you to just explore this individual chart and make any additions or removals that you want without impacting the rest of the of the of the life board. So if you click explore, it pops out of a chart. And if you wanted to do some analysis on this chart, but without creating your own answer or kind of screwing up a liveboard, it allows you to do that. So you can add some filters of your own. You can add some columns if you want. You can replace some columns, and you can even do some comparison with with other columns that aren't in the view. And you can do all this, without actually affecting any of the underlying answers. And once you're done, you can just click off it. So, again, it puts that power into users' hands to do some analysis exploring themselves, without having to go to an analyst or a developer, to ask stuff to be built. And that's kind of yeah. That that's what we'd compare to our traditional dashboards, our live boards, but with few more kind of extra features unique to Forceport. Last thing I'm gonna show is I'm gonna go back to this home page here. And if you noticed at the start, you might see there's also a search bar up here as well. So what I showed previously within this search data panel was the traditional method method of search. It's natural language querying, but you do have to kind of reference the the column names and the the correct keywords to build out the chart you chart you want. So for some people who really don't know the data much and haven't had much training on on FortisBot, they might not be able to get to groups for instance. So around twelve months ago, thirteen months ago, FortisBot introduced a new product called Sage. Now what Sage is is AI integration with a lot large long large language model behind it that allows you to ask a question, in kind of plain English or whatever language, your system's in. What I mean by that is I don't have to actually reference, the names of the columns or the keywords. I can simply kind of write what I want to write in the form I want to write of it. So if I was gonna say something like top selling items this year. Now I don't have a column called selling items, top selling. And I don't have a column called items. I have a column called item types. So if I was to search that in the search data, it wouldn't make any sense because I'm not referencing the correct keywords or the or the or the correct, column names. But if I was to click enter here, it will now generate an AI answer. And what it'll show you is the, the query is actually happening. So from what I've typed in, it's assumed, I mean, top ten by sales for item type for this year, which is exactly what I want. So it it it's assumed that top selling means top sales, which is correct, and it assumes items means item type. I haven't programmed this in. And it's simply learning that's what it's learned from our data. Essentially, that's what it's assumed. And as you can see, it's it's got it pretty much correct as what I want it. So it's got a bar chart showing the top ten items for sales in twenty twenty four. What I could also do is I actually wanna show a column chart and not a bar chart because it's a bit scrunched up in this view. So I'm gonna click column chart, And now I get that view kind of showing vertically rather than horizontally so I can see all the names of, of the products in view. Now I'm sure a lot of you have used LLMs before in for work or for for a hobby. It's very they are obviously a work in progress. So it's not something you should expect perfection for, and it's something that you should really kind of use use cord sparingly, I guess, but it's improving kind of version on version. It's, no doubt that it's the future, and this will this feature is gonna get better kind of with each release. But if you're typing in stuff and you don't think it's it's getting the correct result, you also have this feedback loop that's built into built into Fourspot as well. So if you notice at the bottom of the answer, you have this did we get it right, yes or no buttons. So if they get it right, it's great. And if you click that, it means if someone does a similar search again, false positives knows that this is the correct view it wants from this query. If you got it wrong, you can click that. Thumbs down. And what it does then, if you go to data, you can look at Sage feedback inquiries, and you can see these queries that have been run and whether they've been rated as as good or bad, essentially. So I had a colleague said which item sold the most in January. He didn't think it gave the answer that he wanted, so he fed back. And from there, you can click these dots. And if you click fix the answer, you can actually go to a separate thing and show what you wanted to actually show when someone types in that. So you can kinda train it as you go along on your own data if it's not getting the results you want. But like I said, it's a it's a work in progress, but it's getting more and more powerful, with each each feature. And to demonstrate that, you could say something like, top five. And let's say we typed we spelled products wrong with two s's. And let's say yeah. So products, the field name is products, product name, and we're typing in products, spelled wrong with two s's. So let's just see if it's able to do it. So, again, it's it's got it pretty much spawned. So we've got the top sales. It recognizes recognizes it's by product, and it's got the date of last year. So you can make your spelling mistakes. You can do abbreviations. You can type as if you were in Google, and it's it's gonna generate you a result. And if it's not right, feel free to use that use that feedback loop. But, again, it's bridging that gap even further for the technical for a nontechnical user because they don't even need to be able to reference keywords and columns. It's it's all kind of built in built into this product. Now some of you might have thought about it. I think you can't see this. You don't have this feature enabled. Again, this is still kind of classed as, not not a beta feature, but it's available to be turned on, from the admin panel by your administrator. So you can if you're not confident in your datasets, you don't want people doing these AI queries, you don't you can basically just not turn that on. And I think they are off by default. So don't feel like this is gonna get unleashed on everyone. It can be disabled. And you can also do it as individual group permissions or user permissions as well. So you can have it turned on from admin from a site perspective as an administrator, but you can limit who uses this feature. So that can be useful for if you want some people to sort of test it out. You could you could you could create a group called Sage users, AI users. You can test this feature out, make sure it's working well on your data, and then you can release it to to everyone else after. And what, another concept of what Sage does within ThoughtSpot, so they also have AI built in elsewhere as well. So if you were gonna go to data and go to your worksheets or tables, and then it's good to it's gonna go to sample retail apparel. You also have this concept of synonyms here. I don't have any that reference here. So what synonyms do is that when you're searching for data, it might you might want product to also be called product name or store to be called shop. So when you type in shop in search data, it will also it will know that you're referencing store. So if you have a dataset, that has names that is might have some obvious synonyms, the AI will generate those synonyms as well, so you don't have to. But you can also add them yourself manually as well. So it's kinda got AI working all the way through the product, as well. And the last thing I just wanted to show quickly that I forgot to demonstrate was alerts. So I'm just gonna create a KPI quickly, as we've seen before. And I'm gonna make it a, KPI card. I'm just gonna pin it to my live board as well. So down here, we've got our KPI. And if you notice, a bit different to the other chart is if you hover over your KPI chart, you have this extra kind of alarm, alert button here. So any KPI that you can create in ThoughtSpot, you can automatically create an alert from. And what an alert does is that it can, it can send an email or you can do it through kind of a Slack notification for someone if a certain threshold is hit. So this KPI, if it changes by x percentage, I can add certain subscribers so they can always be alerted when that one number changes. So if your sales have gone down five percent a month, you may want it, you may want to show that to various salespeople or operations people to let them know that number's going down. So you can also you could do a threshold alert, and you can also do a scheduled alert. So maybe you just wanna send that every month, regardless of that threshold, so people know how that number is doing. So, again, it can bring insights to people who don't even need to go into FortSpark. They can it's really about kind of bringing sharing that data to anyone who needs it, whether that's inside or outside the platform. And then any alerts you do set up, you'll be able to monitor up here. We haven't got any setup on this on this demo instance, but any any alerts you do, you can edit, and change any thresholds or schedules from up there as well. So, yeah, that's that's pretty much it. That's a whistle stop tour of ThoughtSpot itself. Like I said, we are a ThoughtSpot partner. And if you have any questions around the platform or or want kind of more of a kind of one to one session, please feel free to let us know. I know a lot of you on this call might be pro clients of ours as well. We also, assist within FortSpa. So if you have any technical questions, that you need some help with, we do cover for Spot as well as other other tools as well.