ThoughtSpot full walkthrough: search, Spotter AI, data modelling and live boards

0:00Okay, yesterday it was Sigma.

0:02Today it's ThoughtSpot.

0:04Um, it's more

0:05detail.

0:05In

0:06in Sigma's video, I said Monday, Tuesday, Wednesday, Thursday.

0:10And then as soon as I edited the video, I realized, oh

0:13crap.

0:14I actually need to get these out sooner.

0:17So I actually

0:18just shipped it the same day, if that makes sense.

0:21So

0:21just just ignore the days.

0:23Uh

0:24the next five days

0:25we're covering different platforms.

0:26I need to get through them by Wednesday for a good reason.

0:29So yeah, that reason will

0:31Will become apparent um later on.

0:33Today we're looking at ThoughtSpot

0:34and um

0:35look, I have to be up front and say something about ThoughtSpot, which is

0:39When I put ThoughtSpot

0:40on my list of tools to

0:42go through to discover, to be honest, I I wasn't expecting much.

0:46I wasn't expecting much at all.

0:47I was expecting to come across the same story that I've um heard before.

0:52That said

0:53Of the four fat

0:54poems that I'm going through the next day, it is

0:56it is by far the one that has impressed me the most.

0:59Partly because I had preconceptions that were incorrect.

1:01but also

1:02because actually

1:03I've

1:04I've found things in it that I think I prefer

1:07in ThoughtSpot compared to anywhere else.

1:09So there are certain capabilities in ThoughtSpot

1:11that I I

1:12that I genuinely think are the best, if that makes sense.

1:15And

1:16but, but, of course, we need to get into the video.

1:18We need to get into what those

1:20capabilities are

1:21this one is also slightly different

1:22because I had a relationship with ThoughtSpot

1:24before this video.

1:25I've had a relationship with Sigma as well for the record.

1:27It's just a very different one.

1:29Thorspot went about it a very different way, and this is sort of one of the uh behind the scenes creator things

1:35where when I reached out

1:36to uh thought spot um their CEO got in touch with me which was incredible um let me just put my phone on silent so that I don't get another message

1:44um

1:45And

1:46um

1:47when the CAO reached out, I also knew a few people who'd gone to Thoughts Spot from

1:50Tableau.

1:51And um yeah, they were just very receptive.

1:54So they put me on a path.

1:55And to the to be honest, I said I wasn't gonna cover the technology unless I knew it and understood it really, really well.

2:01I hate making content about things I don't understand

2:03I've I think I've I've digged at other creators for doing that.

2:06So I wasn't about to repeat that habit myself.

2:09So

2:09when we made this video, um this was the first step, um, Kitane and Francois, uh

2:14CTO and CEO of

2:16the company got on the channel.

2:18I've I've never had that level of exposure, not even from Tableau, which is pretty wild.

2:22A

2:23CTO, chief product officer, sorry, CPO.

2:26Um

2:26And CEO came on the channel and we had a great conversation.

2:29It's had really good viewership, one

2:31views, and and actually the engagement on that is also really, really good.

2:34It's part of a series if you haven't already figured it out.

2:37Um I've kind of been teasing these platforms up front

2:40so it's not um unusual.

2:41And so we won't really dig into

2:43much further.

2:44They unfortunately pretty much um

2:46let's just say they shut down, they

2:47kind of changed ownership and stuff.

2:48stuff like that.

2:49So I won't pursue

2:50Ansa

2:50that much.

2:50But back to ThoughtSpot.

2:52Had a great conversation with them, a one hour video.

2:55Behind the scenes, I was talking to their marketing team.

2:58I actually know someone who works in their marketing team as well.

3:00It's been very, very good relationship to have.

3:02And um

3:03they put me and enabled me to understand the product better.

3:06So they put me through the equivalent of what is called the Thoughts Wat Cloud Essentials three-day in-person training.

3:13Now, uh, this training takes different formats, but the reason I say

3:16they put me through the equivalent of this is because it was more or less everything that you see on this list.

3:20So

3:21um

3:21we spent

3:22uh

3:24Four hours on the end

3:25user experience, which is kind of what I'm gonna touch today a little bit.

3:28We went into the advanced stuff, then we went into the data connectivity.

3:32And I'm gonna say the data connectivity and data modeling really impressed me.

3:36This is the bit where I was like, okay, no, this platform

3:41really

3:41jibes with the way that I think data modeling should be done in a modern analytics

3:46uh platform.

3:47And then

3:48Um there are also other capabilities we didn't cover, but then they also showed us some of the AI capabilities that is infused with this and it generally

3:54worked

3:55the kind of stuff that we've sort of hoped Tableau

3:56would

3:57have

3:57working already um hasn't worked.

3:59So um

3:59as an endorsement

4:01I would say this is a very good way to get to know Thoughts

4:04but

4:04I actually think most companies instead of like hiring consultants

4:07and stuff like that to come and figure this out.

4:09Just send your three most talented people in in the BI tool that you currently have

4:13to a vendor training for three days.

4:15Let them fully immerse and ask lots of questions.

4:17They will ask questions and and this is what I I I sort of understood.

4:22Um

4:22when I saw other people from other businesses asking questions in this

4:25in this training, it was actually really, really good.

4:28It was very um detailed and the way that the

4:31questions were structured sort of led you to really

4:34understand

4:34uh the

4:35use cases in depth.

4:36So and and by the way, everyone in my training was transitioning

4:40from

4:40Tableau more or less.

4:41Not not all of them 100%, but most of them were coming from Tableau to

4:46ThoughtSpot.

4:47And a few of them

4:48were just basically in there to sort of enhance their training.

4:50If that makes sense.

4:51So

4:51a very good training.

4:53I highly endorse it.

4:54The trainer

4:54who does this, I was lucky to be uh trained by Jordan

4:57Suffon, who's

4:58sort of leads the training program.

4:59Excellent training, three days of

5:01uh

5:01in person

5:02training like you gotta know your stuff to be able to do that um

5:05but i will stay on this page

5:07because that training here is advertised at a thousand dollars

5:11so um

5:12to be put through that experience um as part of this process I have to kind of

5:15it's technically not a sponsorship but again you know me I'm trying to be transparent and put these things forward.

5:20Um, I didn't arrive at this perspective without some sort of investment from ThoughtSpot.

5:24And so I have to call that out.

5:26Um, but uh

5:27generally speaking, their customers tend to go on this training anyway, if that makes sense.

5:31So it's sort of a balance between

5:34um

5:35You can just pay for it and go and do it.

5:37And there's also there's also a virtual virtual one which is actually pretty good and they have loads of sessions available.

5:42So

5:42This is what I went through.

5:43I found it really valuable.

5:44And it helps

5:45cement what we covered in this one hour video that I had with ThoughtsBot.

5:49So this is all to say that I

5:51come to you.

5:52with

5:52uh

5:53probably the best prepared perspective

5:55ever

5:55had of a new platform in my entire life.

5:58Um

5:59uh even

5:59even

5:59Tableau

6:00I didn't get to know this well.

6:01So um hats off to Thoughtspot

6:03for putting that much time and investment

6:04in enabling me to hopefully now enable you and

6:08give you an unbiased perspective.

6:10That is what I'm here to do.

6:11Uh there are things in Thoughts

6:12Bot that I find sort of weird and confusing.

6:15But every BI platform has to make some sort of philosophical decision

6:19and some of those philosophical decisions really work well in some scenarios and they work really badly in other scenarios.

6:26The other thing I want to call out is I think in Sigma's video, what I didn't really highlight is look, I am only covering the like um

6:32the I'm gonna say the old school now because it feels like it very much is.

6:36the old school way of building dashboards

6:38because I don't think uh people are going to be like, you know, tinkering

6:41dashboards for that much longer.

6:43But the reason I'm starting from that perspective

6:45is because I need a level playing field for you to understand the mechanic of how those platforms work.

6:51And then

6:52when we start talking about the AI capabilities on top of that,

6:55I truly believe then you'll be able to make some sort of value judgments about

6:59how AI plugs into those things.

7:01And then another thing that BR

7:02platforms have to do is they have to reconfigure themselves.

7:05to uh sort of have this other mode and that's always going to start from this sort of BI platform baseline.

7:11Now the other thing I will say is that Thoughtspot

7:13have this AI capability which is

7:14Generally named around a dog, it's called Spotter.

7:17We'll get into that.

7:18But

7:18all that to say,

7:20yeah, like we're here and I want to start going through ThoughtSpot.

7:23So like all the other videos, I've given you my own preamble

7:26The point of this video, let me just close some tabs so I don't get distracted.

7:30Um

7:31the point of this video is to essentially go through

7:34um

7:34the whole environment.

7:35So what I have done, I have not actually uh set up an environment in ThoughtSpot

7:40because I did it in the training, you're kind of gonna get like um

7:43a real-time playback of how well the training worked on me

7:47Um

7:48it's not a reflection of the trainer, I must add, but nonetheless, um

7:52we're gonna go through this whole experience and you're gonna see the onboarding process as well.

7:56You didn't see this with Sigway

7:58because that was sort of a weird setup I was sort of invited into

8:00environment

8:01So I was just you know

8:02plunked right in as an admin so you don't really get um an

8:05onboarding experience that's like super detailed and the way I had to set it up meant

8:09I kind of didn't have that.

8:11So, um, yeah, what I'm gonna do is just go through the experience.

8:14You're gonna watch it with me if you're doing um a Thoughts What trial for the first time, this is exactly what you see.

8:18So

8:19I have no special treatment here.

8:20I have to do the work myself.

8:22We might even have to set up a snowflake connection, but I might try and use what's already in there just to make the video a little bit more interesting.

8:28The other thing I will say is if you've checked out ThoughtSpot at some point in the last five years,

8:32maybe a long, long time ago, and you you have this idea of what it is

8:36I highly encourage that you take another look and just look at it again.

8:41It has come a long way.

8:42Something I've been saying with all these platforms is the floor has risen very, very high.

8:47and the ceiling hasn't changed that much.

8:49So all BI platforms generally can do the same thing.

8:53That wasn't true three years ago.

8:55It is now.

8:56AI is making it very easy for these platforms to level up their capabilities and also pivot very quickly.

9:02So

9:02let's get into this onboarding.

9:04What department do you work in?

9:05Datra and analytics, no surprise.

9:07What industry?

9:08Ah

9:08man, I always

9:09I always um

9:10I always wonder what to pick here.

9:11I think I'm gonna pick the one I have the most affinity

9:14That's either financial services or retail and CPG.

9:17I'll choose retail and CPG, uh, one

9:19to fifty employees.

9:20We're creating your custom data set.

9:21Please hang in there.

9:22So what they've done there is they've taken um those sort of specific settings and they've

9:27they're kind of loading it up with some um dummy data so here you go

9:31um

9:31this is your

9:33sort of onboarding experience

9:34so

9:35you

9:35got some uh a table here it's pretty straightforward

9:38data analytics data set for you your

9:39industry is ready

9:40choose your use case

9:42And as soon as you do that, you basically

9:45get these sort of three things.

9:47Now, I could just ignore these, but

9:50start with a question your team is already asking.

9:53So

9:53Are we looking at um

9:55retail data pipeline, data quality, SLA messes?

9:59I'll

9:59choose pipeline features and then just say let's go.

10:03And then what this does is it brings you into the main environment.

10:06Now, what it's chosen to do is drop us straight into the AI experience.

10:10And to be honest with you, if you were

10:13um

10:14uh

10:14uh

10:15Uh

10:15a few of the

10:16I'm I'm hold on.

10:17I'm like looking at this at the top and it says

10:1939

10:20days of free

10:21trial.

10:21You know what?

10:22I think some of the I think some of the thoughts

10:24saw me sign up and

10:26sort of changed my trial

10:28That's kind of funny.

10:30Um, okay.

10:31Um

10:32that's uh

10:34I don't think everyone gets that experience.

10:36Uh oh I will add

10:37it says that there's a

10:38uh when you click on free trial it's a 14 day free trial

10:42i think i think

10:43i think their crm is very good that's all i can say

10:47so um

10:48I don't even know how to get rid of this.

10:50It looks really bad.

10:51Let me just, yeah, like how do we how do we

10:53uh

10:54view pricing, book expert one-to-one

10:56help?

10:57Um

10:57Okay, let's just refresh the page

10:59and

11:00no, yeah

11:01I guess it's just gonna stay there.

11:03I just

11:03it

11:03I think it's a trial, so it's just gonna stay there this whole time.

11:06That's that's what it's gonna be.

11:08So that's fine.

11:08Anyway, let's get back to the video.

11:10So it's dropped us in spotter.

11:12Actually, this is not where I'm gonna start.

11:14Okay.

11:14It is actually very good.

11:16Spot

11:16is

11:17very good.

11:17I'm not gonna start there.

11:18What I'm actually gonna do is sort of just walk you through the interface.

11:21So

11:22If I go over here on the left hand side, I'm an administrator, so I see three things.

11:26I see the insights tab.

11:28There is the uh, we're gonna call it the data workspace, which has lots of different data sets

11:34We have uh the developer place, which is where the embedding and API capabilities live, and then you have the administrative space where you can be an administrator and there you can see um

11:45my

11:46email.

11:46I'll need to blur that out.

11:48But

11:49for most um

11:50everyday analysts, this is what you're going to be seeing.

11:53Now

11:54what I will do is

11:56uh sort of go through each of these capabilities just to give you an idea.

11:59And what I want to do is sort of walk you through the

12:02Mechanic of building

12:05I'm gonna say a dashboard, but that's not really what we're trying to do.

12:08So you've got insights, which is broken down into four areas: home, spotter, agent spot, and search data.

12:14Now

12:15Most of

12:16people's experiences really actually start with search data.

12:20So

12:20search data is a very analogous experience to a sheet inside a tableau.

12:25a Canvas

12:25and Sigma.

12:26It has a just a very different philosophy.

12:28And the philosophy is that look, when you go into a data set, you're typically looking for

12:33the answer to a question.

12:35So

12:35search is how you ask that question

12:38and an answer is the result that you see here in the canvas.

12:41That's the best way, the simplest way to explain it.

12:43So if we go in here and we say, okay, um

12:46you can actually do this a couple ways.

12:47You don't have to like

12:49uh

12:49type it so for example um

12:51if I go down

12:52and I've chosen a data set that I actually really do not like

12:55so what we can do is we can go down in here and you can actually change the data that you want to use

13:02You can uh select multiple sources as well, so it's not just a single sort of select capability.

13:07You can ask data or ask questions of multiple data sources

13:10at a time, but just to keep this um, you know

13:14Demo simple.

13:15I'm going to use a retail superstore data set.

13:18Select that just so that we can actually uh ask some uh more interesting questions.

13:22Now

13:22we'll

13:23I

13:23will no, actually if I go back into that, it's not the best of data sets.

13:27Let me see if I should

13:29I should have really um

13:30I should have really uh connected up something of my own.

13:34I think this is fine.

13:35I think this is fine.

13:36I think this is

13:37is fine.

13:38Um

13:38yeah we just we

13:39just use this for now.

13:40So

13:41let's just say I want to see uh the quantity uh purchased in a specific

13:46Okay, so I've chosen those two things.

13:48I'll go over here and hit go.

13:49And what it should do is it'll figure out okay, what's the best way to look at this?

13:53Here's each

13:54city, here's a bar

13:55chart.

13:55And then you get your chart.

13:56It's dead simple

13:57now.

13:58Um, you can change the way this is sorted here on the left-hand side.

14:02You've got some basic uh controls

14:04here for the uh answer that you're looking at.

14:06Think of this as an answer

14:07if that makes sense

14:08Um

14:08and over here on the right

14:10hand side, for each uh of these responses that you get, you've got the ability to change the chart type.

14:16So here are all the different chart types.

14:18You've got the ability to change the way that it's laid out.

14:20Um, you've got the ability to change the specific parameters and formatting that you see

14:25And you can even do things like swap the axis around.

14:28So this right

14:29hand side panel is sort of like your formatting and customization view, and it has a lot of things you can do for this specific chart.

14:36And it's always chart specific

14:38because of course

14:39Right now we're not looking at a dashboard canvas, we're just looking at an answer canvas.

14:43So this is a search, it's creating an answer, and when we pin that answer, when we save that answer, it goes into their equivalent of a dashboard, which is called a live board.

14:52So

14:53for for funsies, let's just um

14:55yeah, let's just let's just say okay this is fine.

14:57I'm quite happy with this.

14:59We'll close that by clicking on in again.

15:01And um

15:02you can look at this um in the Tableau

15:03V

15:04if I just look at these options here.

15:05So it's got the the information that you'd look at there.

15:07You can look at

15:07as a chart or as a table.

15:09And there's a couple of different things you can do.

15:11Now, one thing I'm gonna call out already that I really, really like

15:15is that pretty much the whole ThoughtSpot

15:18platform uses something called a TML structure

15:20Okay, so you can export um

15:23the TML uh of this file.

15:25I'm just gonna put it on my desktop.

15:27Okay, what blows me away is that

15:29It it is essentially a YAML

15:31instantiation

15:32of the asset that you've just built.

15:34So if I put this on the right

15:35hand side, um, let's just do this just very quickly.

15:39Let's get this

15:39uh

15:40like that.

15:40So if you put the if I put the chart on the

15:42uh thing on the right there, if I put them side by side, what you can essentially see is you've got your answer.

15:46I haven't obviously given it a name.

15:48Um

15:48it's got this data source, this is the query.

15:51A

15:52as someone who

15:53um

15:54Who thinks about data portability

15:56and who thinks about um

15:58semantics

15:59and metadata?

16:00This

16:01this to me

16:02is absolute gold.

16:04Because

16:05It's very easy to take this uh

16:07format

16:08and move it to

16:10whatever other analytics

16:11platform you want to go to

16:13because

16:13It

16:14essentially

16:14describing exactly what you're looking at.

16:17And you don't even have to choose necessarily a

16:19bar

16:20chart, you could choose something else.

16:22The point is

16:23the way that this is drawn is essentially codified here and it's quite specific to ThoughtSpot, but then you can also translate it to another platform.

16:30So that's something that you can

16:33You can look at uh

16:34in your own time and it virtually exists for everything

16:37connections, the whole entire platform.

16:39So

16:39like

16:40that's a bit of a side quest.

16:41I don't know

16:41why I'm going into YAML at this point of a tutorial for a newbie

16:46But uh it's something that I love and I was like, I've got to show it to you while I've seen it before I forget.

16:51So there we go.

16:52Uh what we're gonna do is we're just gonna save this and we're gonna yeah, it's just given us a name, Total Quantity

16:56Search.

16:57Purchased by City.

16:58Let's save the answer.

16:58So you did a search, it

17:00saved the answer.

17:01There you go.

17:02That's pretty much good to go.

17:03Okay.

17:04Now

17:05um

17:06We can actually go back uh to search data again.

17:08Now the reason I've done it like that um

17:11is

17:11I could have actually cleared the sheet and then gone and saved a new one, but the reason I've done it like that is because I didn't want to overwrite the existing one

17:17And I've forgotten already how to just basically, you know, just start a new search just without having to do that.

17:24So I've just done it quickly for the purpose of the video.

17:26Again, um

17:26I don't spend

17:27I've well I've never used ThoughtSpot

17:29in a

17:29client settings so that is a mechanic I'm still learning

17:32so let's go ahead and uh let's look at something different let's look at uh product

17:36region

17:37uh

17:37sales and

17:38Let's just

17:38hit it go.

17:39We'll bring in a date um in a minute.

17:41And this is a bit more of an interesting, I guess, chart.

17:45Now what you'll see is because I've put product in front of region.

17:48It's gone and used that granularity to direct how things work.

17:51So

17:52what you could do is you could say, okay, I'm actually gonna get rid of that and I'm just gonna type in here

17:57region.

17:58Okay, and you'll see that it redraws the chart.

18:01This time it's drawn a pie chart, and then we can then go and say

18:03okay, product, and um

18:06it automatically does that and boom, now it's come back

18:09now

18:10The reason it's done it this way is because the first time we drew it

18:14um

18:14I think it came back with

18:16product.

18:16So

18:17it's actually a really interesting question.

18:18Let me just let me just clear this.

18:20Let's just go sales.

18:21Let's just go region

18:23Okay, and then we go product.

18:26Okay.

18:26Does it do the same thing?

18:28Uh

18:28it does do that same thing.

18:30Okay.

18:31Um

18:32Um

18:33if I choose

18:34by

18:35uh

18:35and

18:36then

18:36uh

18:37region

18:37uh

18:38by

18:38product.

18:38It's got these um

18:40keywords

18:41Which is

18:42which is sort of an interesting

18:44capability.

18:45And

18:45just just as I type this, you can see that it's got like a whole bunch of them.

18:49Some of them are filters, some of them are

18:51um

18:52you know, just terms in our field

18:54and the rest of them

18:56are

18:56um, you know, these are all mostly filters, but there are, for example,

19:00um

19:01special terms that can do specific things and help you describe specific aggregations in the whole entire process.

19:09So again

19:10I don't want to go off on a tangent and go and sort of Google and search those because I think that's a little bit much for this point in the lesson.

19:16But what I'll do, what I'll do is let's just go over here on the right-hand side.

19:20Let's choose a

19:22more

19:22interesting chart.

19:24What let's why don't we do this?

19:26Let's just do region

19:28uh and then we'll do a quarter

19:30uh

19:31quarter

19:31of a year.

19:32Boom.

19:33Okay, and then

19:35yeah, it also gives us a line chart, and then we can actually click on this and just sue, let's do

19:43um

19:43month

19:44of year actually

19:45so let's see

19:45sales

19:46month

19:47uh

19:47of year uh i think it's that one

19:50uh

19:51practice

19:52month of year and then what else

19:54are you looking at

19:55region

19:55This

19:56the different regions.

19:57There we go.

19:58Now we can hit enter.

19:59So we should get uh

20:00line charts for the different months of the year.

20:03Okay, fine.

20:03All right.

20:04So I've gotten what I sort of could imagine

20:06my head.

20:07And

20:08what I'll call out

20:09actually the philosophical thing here is really important.

20:11Um

20:12the way

20:13that

20:13this

20:14search experience work

20:16My muscle memory isn't quite sort of tuned to it.

20:19In Sigma and in Tableau, they had the philosophy of the drag and drop.

20:25Now

20:25strictly speaking you can tick things, but if I if I try and drag and drop, like that's not that's not happening.

20:31This is not what this interface is designed for.

20:33So

20:34this is what I mean by philosophical um

20:36decisions.

20:37And it's not a dig really against the platform because if you haven't used anything else and this is what you do

20:41use daily, you wouldn't have this sort of intuition to want to drag and drop.

20:46And some people will argue, hey, drag and drop is the best thing.

20:49But it's a bit like saying, um, hey, um, I prefer running to cycling

20:54Okay, uh

20:54and you're someone who's never s

20:56you know, really cycled properly in their life.

20:58Like you don't really know what you're missing out on until you really get into the sport.

21:01So um I'm not gonna dig on that

21:04just because, you know, I'm not used to it, but I will call out that it is

21:08it takes a little bit more cognitive load just to understand how to kind of get this search interface working.

21:14That said, I don't think new users of Thoughtspot

21:17today will use this

21:18experience that much.

21:19I think they'll just go into Spotter, describe exactly what they want and get it the first time.

21:24And I'll show you that experience in a minute.

21:26In fact throughout the whole training we basically spent

21:28Uh the

21:29best part of like 20% of it

21:30learning how to do it the old school way and then in about 20 minutes, um

21:35like literally like minutes

21:37just doing it with AI immediately and getting the right answer every single time.

21:41time

21:41so it was pretty good.

21:42So anyway, let's let's just uh save this.

21:44Um I'm not gonna pin this

21:46just yet because we don't actually

21:47we haven't actually created our live board

21:49so I'll just call this

21:50total sales by month of year day.

21:52I'm not gonna bother with the with the with the titles

21:54um

21:55and we'll just leave that as is okay

21:56uh so yep that has saved that

21:58so let's go ahead and search data again

22:02now

22:03there is actually another way to get to this if you go here and hit

22:05Plus

22:06you can um create an answer, create a live board.

22:08So this is almost exactly the same thing.

22:10So creating an answer kind of just brings you back to this uh place as well.

22:14So the last thing I want to do is

22:16we have some

22:18um

22:19information about state.

22:20So let's do cells

22:22by

22:23state.

22:24And I'll try not to use

22:26my mouse.

22:27I'll just sort of type as we go.

22:28There we go.

22:29We've got a

22:30a nice map let's just go ahead and save that and then there's our map so we've created three charts we've created a bar chart

22:36a line chart a map

22:38uh

22:38I guess for posterity

22:40we should go and um

22:42create a pie

22:43chart

22:44I'll do it with sales again.

22:46And by the way, when you click on these, it does give you a little bit of context.

22:49It's actually kind of really useful to know

22:52It gives you a little bit of metadata that's actually quite useful just to be aware of

22:56and you

22:57it gives you answers as well, um, things that other people have already done, if that makes sense.

23:02So

23:02Um you can kind of get those and if you just click on those

23:05it kind of deep links

23:06you into that answer

23:08and you can go off and do uh the chart just by clicking on that.

23:11So this is it's quite nice to have these sort of pre-prepared things that you can ask of the same data set.

23:16of the same data model just ready

23:18ready to go.

23:19So I I'll go ahead and pin that.

23:21Not that I was actually asking for this.

23:22Uh

23:22I think this is because I didn't create it.

23:24So let's just hit go on that.

23:25Again.

23:26What what went wrong?

23:27Something went wrong.

23:28Thought Spot, if you're watching this, um, what is that error?

23:30So

23:31um

23:31let me do this.

23:33Uh

23:33let's do this.

23:34Last year

23:35versus

23:35this

23:35year.

23:35It's doing a comparison, so maybe that's the

23:38Let's just try and save this.

23:39I don't even care if the chart is good.

23:41Um

23:42it's a book expert one-to-one

23:44help.

23:44Something's something's not right here, so we'll we'll just we'll just put that to one side.

23:48I'll I'll I'll let ThoughtSpot

23:50uh

23:50off

23:50on this one.

23:51I'll just go back to creating things uh

23:53how I wanted to create them.

23:55Maybe there's an issue on my side.

23:56So let's just go ahead and put sales

23:58and

23:58uh we'll just get a uh a total number here and what we'll do is maybe let's bring in

24:04uh item type into this

24:05how many item types are there

24:08Um

24:08let's see.

24:09Okay, that's that's

24:10we can we can maybe get away with a donut chart for that.

24:13I think

24:14it might be a bit busy, but it will work.

24:16So let's go ahead and save.

24:18So

24:18we've got um

24:19several answers that we've created, and if you want to go and see them, um

24:23we can go here to answers.

24:25So you search.

24:26Um

24:26to create an answer.

24:28Once you have your answers and you've saved them, they live here.

24:31Surprise, surprise.

24:32So the administrator created some of these for us.

24:35And then these ones I've created

24:37myself.

24:37So this is the one that I created that seemed to have an error.

24:40It must be because I've got this environment

24:42um from ThoughtSpot

24:43and so it's just brought that in and it's sort of made it um

24:46it's

24:47sort of made it into the environment but maybe my security isn't right.

24:50I think when I set up my own snowflake and everything this stuff will work a lot better.

24:54So

24:55I've got all these.

24:56You

24:56probably asked, well, how do I put them on a dashboard?

24:58Well, this is where we go to their equivalent, which is called a live board.

25:01And a live board

25:03lets you essentially build these

25:05um

25:05capabilities in.

25:06Now

25:07you've got uh AI capabilities and spotterviz

25:10is their ability to build dashboards.

25:12You've also got the ability to just go in here and uh do things yourself.

25:17So I have to say

25:18When I did the training, this wasn't the experience.

25:21So I think they've definitely updated it in there.

25:23But at the bottom here, you've got this option to add an answer directly.

25:27So I will come back

25:28to

25:29uh the experience.

25:31that we just saw in a minute.

25:33Um and what it's done here is it's taken me back to the answers.

25:35So I now sort of understand what it wants me to do.

25:38Go back to the answer.

25:40And I think in here

25:42I can um

25:43well I'll just have to go to each one of them.

25:45Let's add to a live board.

25:47Where are we?

25:47Where are we?

25:48Where are we?

25:48We where are we?

25:49Okay, let's go back.

25:51Let's

25:51create a live

25:52board.

25:53Let's give it a name

25:55first.

25:55I forgot to do this, so

25:57and

25:57like

25:57board.

25:58Um, this is a really good example like philosophical differences

26:01because

26:02the the

26:03the mechanic of where we go and how we go to do this

26:06is a little bit interesting.

26:08Now

26:08it's desperate to get me to just

26:11Ask

26:12AI, which is I guess a really good experience, which is why they're desperate to do it.

26:16But I just want to show you the non-AY

26:18AI way to do this.

26:19So Tim's Liveboard is there.

26:20Okay.

26:21Let me go back to

26:22okay, there we go.

26:23Um

26:24we've got Tim's

26:25this is just a starter page.

26:26So that's fine.

26:27Let's go back to my answers, sorry, and let's do this.

26:30Let's um

26:32I don't want to add it to a collection.

26:34Why can't I pin it to a live board?

26:36Why am I missing the trick?

26:38here

26:38give me one moment

26:39okay

26:40so because i didn't create a live board previously the pin option didn't show what i wanted to to see but anyway you've got the option here

26:46the previous one i created was also there

26:48So

26:49uh we can pin this to a live board.

26:50So let's go ahead and pin that.

26:52So really I should have created the live board

26:54first

26:54and then uh it would have got there.

26:56So if I go to it to announcer, let's

26:58Maybe I'll just open these all in new tabs and

27:01save myself a lot of hassle.

27:03So

27:03here they all are.

27:05You can see that it's already got my live board

27:08sort of

27:09uh

27:09tuned in there so once you create the live board

27:12first

27:12it is actually quite easy to um

27:15oh

27:16this is the 14 day trial

27:17that's

27:17that's not 14 days.

27:19So I think I've pinned all of those.

27:20Um

27:21those have all worked.

27:22So

27:22now

27:23what we can do, let me close some of these tabs as well, just so that you

27:26You can see what's going on.

27:27So let's go back to my live board and there you go.

27:30Those are all the

27:31things that I've pinned.

27:33One of the things you can now do, we need to kind of edit this a little bit.

27:36You can sort of change the way this works.

27:38We can go ahead

27:40Let's just put this over here.

27:42It's very similar to Sigma in terms of the way the layout system works.

27:46I

27:46again

27:47I

27:47um

27:48It is quite embarrassing

27:50to some respects that, you know, Tableau

27:53have

27:53you know, being the leader in the market

27:55somehow has the most

27:58uh

27:59behind layout system

28:02of all the tools that I've experienced

28:05by simple fact that they haven't ever changed it.

28:07in the many years they've been around.

28:09And it it's such

28:10it's it's sort of really frustrating.

28:12The m

28:12the thing that

28:13most developers spend their most time doing is actually the most behind in something like Tableau.

28:18Um

28:18but

28:19nevertheless, um this is also not perfect.

28:22Like this you know the this this this layout dashboard system has its own uh challenges.

28:27Now from a styling perspective

28:28You can uh I don't know, you can change the color, the background, all that stuff.

28:32What I will say is ThoughtSpot

28:33really tries to focus on, hey, listen, look, we've just done you the chart.

28:37Like

28:37Just go with the flow.

28:39And of course, all of these can be customized.

28:41You saw that previously in the uh answer page you could customize each of these.

28:46But the thing that I find um somewhat different is that usually you see if I click edit, it opens the answer and then I can edit it in here

28:53Um

28:54I can go

28:54change

28:54all the formatting if I hit save.

28:56As soon as I hit save, it takes me back to the dashboard.

28:59So that is a mechanic to allow you to change these things here.

29:02But

29:02You see that view, you kind of ch

29:04you kind of edit it in its own context

29:06and you come back, look at it, how it's going, you edit it again, go back in its own context.

29:10So

29:11you don't have this easy way of

29:14Seeing the whole thing in context, right?

29:16So that's a really uh

29:18important thing.

29:19Now, another thing that you might be wondering is well, now how do I uh let me just save this dashboard?

29:24Sorry.

29:24Um, how do I interact with this, right?

29:26So

29:27you've got the

29:28all this um

29:29capability here, you've got all of these things that are sort of separate.

29:33How do you interact with this?

29:34How do you get all of these um

29:35talks?

29:36Well,

29:36so

29:37we want to um

29:38make this interactive.

29:39Um at the moment um

29:41when I click on all of these, it's it's not really doing anything.

29:44The way to do that is to go up here.

29:45So at the top you can add specific things.

29:47You can add a note, you can add

29:49insight

29:50tile,

29:51add a filter.

29:52So that's exactly what we're going to do.

29:54And

29:54we are going to

29:56add a state filter.

29:58And we'll say that it's uh multi-select, go ahead and hit apply.

30:03And

30:03as soon as we do that, where is our filter gone?

30:06There it is.

30:07You can uh you can

30:08can see here that you can go ahead and for example

30:10choose Michigan and hit apply and then the whole thing changes so your filters sort of live here across the top

30:16you can go ahead

30:17and another one if you wanted to so let's go ahead and add region

30:21hit apply

30:22And then because we've selected Michigan,

30:24there's only really one

30:26uh

30:27area available to us here.

30:28But if you were to go ahead and select Indiana, Iowa, Minnesota.

30:33Missouri, uh

30:34you'd see that this uh

30:36drop

30:36down

30:37is still Midwest.

30:38I

30:38did an expert job of selecting

30:41Oh, because Midwest is selected,

30:44I only could select the things from the same state.

30:47So

30:48let's go and select a few more,

30:52hit apply.

30:52I think these are still

30:54um

30:55these

30:55are still in

30:57uh

30:57there we go

30:58we can see all of that there so the selection now gives you all of the things that you want so let's go

31:04hit

31:04apply.

31:04So now when we hit save, now we're in the dashboard experience.

31:09Uh

31:09you can see the user gets this sort of capability to apply filter and

31:13It's

31:13it's fairly straightforward.

31:15It is possible to add interactivity.

31:16I'm just I'm just not sort of dialing in that capability off the top of my head right now.

31:22The thing I'm trying to race to is to show you, well, okay, this is this is the live board.

31:27Um, and it's

31:28I didn't really have a clear idea of what exactly I was trying to show you.

31:31I'm just trying to show you how to put stuff on the dashboard

31:34What does the AI capability look like?

31:36What would it look like if instead

31:39you got AI to sort of do this stuff?

31:41So what I'm actually going to do is hit

31:44cancel.

31:45discard

31:45the changes

31:46and what we're gonna do is go back to

31:49live

31:50boards

31:51and we're just gonna create a new one

31:53create a new live board

31:54and here

31:54you could see that it's got nothing available so I could say okay

32:00One of the things I need to do

32:02is sort of contextualize

32:05the question I'm trying to answer.

32:06Sandy

32:07is a very handy tool because what it does is it's a push-to-talk mechanism

32:11And as I talk, I can literally transcribe what I'm saying.

32:14So let me let me try this again.

32:16I need to analyze

32:17sales, uh, retail sales specifically.

32:20Uh I want to be able to see sales by region and I want to be able to see

32:24the worst performing products

32:25As soon as I let go of the shortcut, it types that out.

32:28And so you kind of didn't see it there, but it was happening behind the scenes and it it does it in real time on device, no sending to the cloud.

32:34It's uh very secure.

32:36It uses on-device capabilities on Macs.

32:38Is

32:38also

32:39available on Windows and it can do on device capabilities there.

32:42So as soon as I send that um

32:44question to Spotter.

32:46It kind of tries to understand the prompt.

32:48Now the prompt is a very loosey-goosey prompt.

32:50It's not like one of the most um

32:52engaging ones, but it does

32:54uh suggests a data source and it's basically gonna take the retail apparel

32:59one it's of high relevance

33:01so it's saying hey which data source should I choose

33:04uh and I'm gonna say uh

33:05one

33:06so I'll just give it number one

33:08so it understands

33:09okay

33:10that's the one I'm gonna use

33:11and then what it will do is it will then now contextualize itself with that data source

33:17and it

33:18'll go ahead and plan some answers.

33:21And

33:21I will say, you know, something

33:24one critique I've always had of Tableau

33:27and

33:28Um

33:28so if you notice

33:29a little bit of Sigma

33:30is that a lot of the capabilities they have

33:33sometimes they're so bleeding edge, but it's so

33:36opaque.

33:37and difficult to figure out

33:39how to get them.

33:41So you have to know your role, you then you have to know if you're

33:43creator, author, viewer, and

33:46once you know that, you have to know

33:48do

33:48does your enterprise license have the right level to have this?

33:52It just gets very, very complex.

33:54Um, what I like about

33:55at least the ThoughtSpot

33:56experience is that once you're in, you just have this capability.

33:59And

34:00Um I asked this question in the training.

34:02It it is my understanding that all of this stuff doesn't uh get charged back to you in in the form of credits.

34:07This is just part of the experience.

34:09At least it is for now.

34:10It might have changed, it might change in the future, who knows?

34:13But

34:14yeah, this is this is just how it is.

34:16So here's a proposed plan for your retail sales dashboard.

34:19Yeah, go for it.

34:21Go for it.

34:22I just wanted to build something.

34:23Um, I like the plan mode that it's kind of in.

34:25It's sort of prompting your way through what it what it was doing.

34:28And so

34:30If you bear in mind before I had to go find the the answers, build them up.

34:35You have to get into the habit of doing those pretty well.

34:38Then once you save those answers, you put them into a live board.

34:41There is some work behind the scenes of the data

34:43engineer or the

34:44analytics

34:45engineer who setting up the data source needs to do to the data model to enable it to be able to do those things well.

34:50I'll cover a little bit of that in uh

34:52in some of the in some of this a little bit later.

34:55But

34:55um

34:56Yeah, this is

34:57this is just going to go away and do the work.

34:59And I have to say, because of the way that the data models are set up in ThoughtSpot, this is actually a very

35:05reliable experience.

35:07You can see some of the answers it's coming back with.

35:09It's actually telling you how it built each of the answers.

35:12So it's it's sort of telling you, okay, this is how I'm gonna compose that.

35:16You don't have to look at this by the way.

35:18um

35:18it's going to uh do this behind the scenes so I don't have to sit here and watch this

35:23but I find it quite useful to see each step

35:27um

35:27because you'll see that it it kind of does it as code and then as soon as the code starts to happen the dashboard starts to come alive

35:34And then you start to see things working.

35:36So at the moment, um, it's building things uh out of view as well.

35:39If I just actually minimize that and just just let it sort of

35:43Let it do its thing.

35:45Um

35:45you can see that it's building, it's still working and it's still uh executing this.

35:49So um yeah, we'll give it it's uh a bit of time, let it finish, and you'll see that it basically did

35:55what

35:55I was doing from the start in a much much faster way.

35:59And it was actually able to um

36:01get me a dashboard very, very quickly.

36:04Probably one of the fastest experience in

36:06building a dashboard.

36:07Now

36:07Sigma has a similar capability.

36:10It does let you do

36:11go into

36:12plan mode.

36:12In fact, it just released a plan mode in the last few months, so you can essentially

36:16plan

36:17out your dashboard.

36:17So I have to say, you know, Thoughts

36:19What got to that feature before Sigma.

36:21I have to say with Tableau

36:22,

36:23I'm sure the feature exists in some part of the Tableau

36:26ecosystem, but it's probably not the one uh it's probably not in the experience I want to use.

36:31Um

36:32there is a

36:33there is an agentic

36:34capability inside of desktop, but it

36:37it depends it depends which

36:38demos and which experience you've watched.

36:41If you've been on like a keynote conference rage, yes, there is this sort of ability to let Claude go out and build out your dashboards

36:48If you're in sort of just a normal everyday desktop experience, you can kind of string it together.

36:53You can even just honestly use Clawed Computer Vision to get

36:57build the dashboard for you.

36:58Uh and there you go.

36:59Anyway, it's done actually done some stuff.

37:01It's created some answers and

37:02it's updated the live board.

37:03So let's let's have a look at the live board.

37:05This is it.

37:06Um

37:06It's still

37:07uh pulling some information uh up here.

37:09So I think it's basically crafting sort of the narrative story for the top of this.

37:13It's giving me a nice

37:14timeline series.

37:15There you go.

37:16So it

37:16crafted this sort of um story here at the top.

37:19And there's a little uh AI icon here in the right.

37:21And we've got a dashboard.

37:23It's pretty straightforward.

37:24It's broken this down in a fairly strange way, but it's it's trying to do it for each of these

37:28I would have maybe chosen this

37:30somewhat different um

37:31orientation.

37:32But yeah, this this pretty much um does the job.

37:34So that's the uh you know

37:36spot of

37:36viz.

37:37If I just save this, um, obviously it it gives me sort of this like walkthrough, hey, you've done this part of the experience.

37:43Now if we go to spotter

37:46The thing it actually landed us on.

37:47What spotter

37:48essentially is, is the exact same experience

37:50but for building answers on their own.

37:52So

37:53if I was to change this data model and I went back to my Apparel

37:57and selected it and I

37:58said

37:59start a new chat.

37:59It starts a new chat because it's technically a different data source and I just asked it

38:04what's going on with my sales

38:06It's actually very uh good.

38:08Again, the data model in ThoughtSpot, um, like just

38:11just take it from me as someone who's actually been to the training.

38:14It's good.

38:14I I always want to give it the trust me

38:16bro.

38:17uh the data modeling capabilities in this tool are really really good and as a result

38:22when you go in and you ask these kinds of questions

38:25it comes back in a fairly deterministic way.

38:27And you can throw it some pretty nasty language and it will sort of come back.

38:30I don't mean rude language.

38:31I just mean

38:33As long as you've done the ro

38:34the right work to set up um

38:36semantics and not not

38:38not semantic modeling, I'm talking about semantics as in how your organization names certain things.

38:43Um

38:43you can pretty much get it to sort of handle those um

38:46all the time.

38:47So you can see here, here's a snapshot of what's going on with your cells.

38:50So I asked it, what's going with my sales?

38:51You got

38:51sales overview.

38:52uh

38:53something happened in February and

38:54March onwards

38:55obviously very quiet November to December so we must be a very seasonal business and I can go ahead and change this to let's say

39:0224 months if I wanted to

39:04So just by clicking in on the answer, hit apply, and you'll see that yes, we are indeed a seasonal business.

39:09So it's very easy for someone to just target the thing to ask the question and get the you know

39:14the answer to what they're thinking.

39:16If I like this, I can pin it.

39:18I can go ahead and add it to my LAD live board.

39:20So now you kind of may be seeing how these

39:23start to come together.

39:24Sales

39:24by region, this is sort of breaking that down for me.

39:27And it's also given me a table, and then it's also giving me items, and it's also giving me top performers in key takeaways.

39:33Honestly, out of the box

39:35I just simply asked what's going on with my sales and it gave me all of these things and each one of these things

39:40by the way if you want to go into them and play in that search experience

39:42experience.

39:43You absolutely can.

39:44You've got the kind of ability to

39:45do that.

39:46You've also got the ability to edit them.

39:48So editing the answer just opens up that answer sort of interface, which is why I started the video in that place.

39:54Because I kind of had to show you where the base

39:56mechanic starts before all of these things start to actually sort of make sense.

40:00Uh and part of the point of these videos is hey, if you've been thrown

40:03thought spot for a weekend and you're being asked to sort of figure it out

40:06This video is supposed to sort of help you do that.

40:08I am by no means finished.

40:10We we've still got a lot more time.

40:12Because I went through this platform in a lot more detail

40:14um

40:15this this is sort of how um it goes.

40:17So anyway, I think I have more or less

40:19um

40:20gone through everything I want to from the

40:22spotter and um

40:23the uh live board experience the spotter

40:26viz capabilities

40:27all of those

40:28are

40:28part of the standard experience.

40:30And um

40:31there is so much under the hood here.

40:33Every single sort of use case you'd ever think of in a business intelligence context.

40:37I think it has been thought through very well.

40:39So this is the insights

40:40tab.

40:41We've just finished

40:42one part

40:43of this video.

40:44This is the insights tab.

40:45Okay.

40:45This is the bit where I was like, okay, now we're talking.

40:49Now we are talking

40:51The data capability

40:53in

40:54this

40:54um

40:55in this tool is

40:56I

40:57actually generally, generally, generally

40:58impressive.

40:59I have no connections, uh, but if you were to go ahead and create one, this is what it looks like.

41:03You'd choose your instance.

41:04In this case, I'd probably go with Snowflake.

41:06You can upload CSVs and Google Sheets and stuff like that.

41:09What it does is it then uploads those um and then you're pretty much off to races.

41:13So if you're to click next, you get the standard configuration window.

41:16You set these up, you bring in um

41:18once you've set up your connections, you can use those connections to do different things.

41:22I will probably do that in

41:23another video.

41:24But if I go to my data objects, they've already loaded us up some data objects which we can use.

41:29So

41:29in here you'll see lots of different uh data objects.

41:33Some of these are demo connections, so you can see

41:35they all come from

41:37different sort of domains.

41:39Along the top you'll see that some of these are models,

41:42some of these are tables, and the nuance is

41:45very subtle.

41:46If you look at this, this is a table.

41:48If you look at that, it's got the uh sort of a table icon

41:51and um

41:52the icon

41:53to the left.

41:53So

41:54It's actually a model.

41:55Now the models, let's go have a look at this model, the hockey

41:58store, uh, is essentially going to allow you to bring in uh

42:02multiple rows.

42:04from

42:04uh different, let's say, tables.

42:07But if I go to the joins

42:08space, you'll actually get to see how that data model is actually built out.

42:12And

42:13This is where I was like, okay, this is actually a very, very good capability.

42:18Because not only does it really let you build these manually, but they also have a spotter capability.

42:24that genuinely automatically build

42:26some of the base

42:27best data models that I've I've

42:29used um ever.

42:30Like

42:31some of the stuff that we were talking about with Kirk in the data modeling uh capabilities.

42:36Um

42:36this

42:37experience was able to generate the equivalent of those sort of complex data models in the training that I was doing.

42:43So at some point I think I'm gonna have to get Kirk back up.

42:46Kirk back on here.

42:47We're gonna try and build out some of the data models.

42:50He can put it to the test and uh we can be the

42:52we can be the judge of whether Thorspot can handle it.

42:56Kirk, if you're watching, I know you are.

42:58Let's schedule another video on Thoughts

43:00Bot at some point in the next uh two to three weeks.

43:03Data samples, uh this is this is a pretty uh straightforward.

43:07You could just

43:07see some of the data

43:09dependence

43:10uh these are the things that are linked to it so it's basically being used in a live board and that's pretty much it

43:16Now if I go back to this, I can't actually edit this model because I didn't build it, so it's not really showing me much, but this is a

43:22I

43:23like I'm honestly

43:24I've honestly been really, really impressed with this.

43:26If I go back, let's try and find a

43:28another

43:29um

43:29Again, because I haven't built these, it's not really going to sing and dance the way I wanted it to.

43:35This is the

43:36um

43:36okay, this is just

43:38okay, fine.

43:38This is not going to show me what I want.

43:40Let's try and find a different data model that we can

43:44can look at see if there's anyone

43:46from

43:47spotter

43:47analytics

43:48time

43:49series this is this looks like a pretty good one

43:51joins with this

43:52model

43:52joins from this

43:53model

43:53okay now

43:54um

43:55I was actually quite lucky with the one I picked straight out of the gate, right?

43:59One is a

44:00uh

44:00entitlement.

44:02All the rest of these are table.

44:04I really need to just go out and build my own data model, I think.

44:07Um

44:08this is a service

44:09now

44:09uh

44:10data set.

44:11So

44:11I mean yeah, this is about as complex as I

44:13probably probably show you and you could see um

44:16how this is sort of set out.

44:18It's just it's

44:19it's

44:20The diagram doesn't really sort of do it justice, um, but the ability to

44:26be

44:27um

44:27This

44:28precise.

44:28And actually

44:29what I will say is the mechanic of how you set up these data models is actually really important.

44:34They set

44:35the permissioning model

44:36so that the person who creates the connection

44:39doesn't have to be the same as the person who then comes in and builds the model.

44:42And again

44:43doesn't have to be the same as the person who comes in and builds the chart.

44:46So you can create some really good

44:48um

44:49let's say

44:50governance in the way that you build

44:52test

44:53and deploy your data sets in here and

44:55because everything builds on top of each other.

44:56So the person who creates a connection can give the connection uh to

45:00can give access the connection to a group of people.

45:02The group of people have different capabilities to be able to touch on that

45:06Those people can then create models and they can create models

45:10which are then given access to a specific group.

45:12But that group doesn't have to be the same group that was given access to create the model in the first place.

45:17So

45:17You have all these um

45:19really nice

45:20nested sort of um

45:22capabilities inside of the

45:24uh

45:24data modeling experiment.

45:25Experience.

45:25Alright, so

45:26let's just look at a boring table on its own, just so you can see what you get in a standard view.

45:31Obviously, all these attributes are grayed out again because they're just

45:34sample data set.

45:35I think I really have to go ahead and create my

45:37uh

45:38my own information.

45:39Actually this has got some um

45:41uh

45:42information that's been sort of de-identified I don't know if that's done in Snowflake or not

45:46Um but yeah, needless to say, you kind of get the idea.

45:49Um

45:50what else have we got here?

45:51So we've got connections, we haven't set any up.

45:53Analyst

45:53Studio is

45:54um

45:55uh

45:55I think it's an experience, it's technically an older experience for building out and doing data prep.

46:01To be honest, is probably something that's going to go away, but we can we can chat about that some other time.

46:05Now you've got sinks.

46:07Um

46:07we maybe saw this before.

46:10Let's go back to

46:11I'll just I'll just

46:12literally give you an example

46:14Let's let's look at retail sales analysis.

46:17So when I click on this, uh

46:19this chart here, you'll see there's an option to sync, and you can see I have the option to sync to sheets.

46:24And so what it does is it says, hey, can

46:25you want to sign into

46:26Google?

46:27And it's basically saying

46:29on a

46:29on a schedule, it can essentially sync this to Google Sheets.

46:34Um

46:34And you can sync it on save or sync on a schedule.

46:37So every so often this kind of data can go off to Google Sheets, which is kind of nice because you're just using this raw

46:42table as your um sort of insight.

46:44Which is um yeah, pretty pretty pretty nice.

46:47What else do I need to show you?

46:48What else do I need to show you?

46:50I think for the start, just to just give you a whistle stop

46:54tour of ThoughtSpot

46:55and sort of where everything is, we should really connect.

46:59uh some

46:59some stuff

47:00so at this point I'm gonna sort of stop going through the experience a little bit there's so much more to do

47:05the last thing I want to show you is actually something

47:08I'm very fond of

47:09I really liked actually about this.

47:11I'm not sure this particular view is going to tell us much, but um

47:15there is a bit um a capability here which allows you to look at the query now.

47:19This is not original

47:21um

47:22Tableau

47:23for

47:23Power

47:23BI.

47:24I think every tool lets you look at the query.

47:26You can even look at the SQL, for example, and see

47:29how

47:29that was sent.

47:30You can even go get that from Snowflake if it's sitting on top of it.

47:33What I actually really liked myself was this

47:35this query visualizer

47:37and this is this is particularly useful.

47:40when you get into really complex scenarios where you need to be able to uh

47:44dissect exactly what's going on with the query, especially when you've got pre-aggregation happening in a one query.

47:50before it then goes off to another query and just being able to see that up front is super super handy.

47:55So I will call that out again as a capability that I think you should be aware of.

47:59The query visualizer is

48:01Top notch.

48:02Okay.

48:02So yeah.

48:03Um

48:04there's so much more.

48:05Um

48:06and I haven't showed you collections, um, answers

48:10we've touched live boards, we've touched, search data we've touched.

48:13uh

48:13Agent Spot.

48:14Um

48:15AgentSpot is actually a separate capability you'll see that it it chucks us out of that.

48:20It's showing in the

48:21in the main experience at the moment, but um I've actually got access to this.

48:25I won't cover it in this video because it it's a very, very new experience.

48:29It is likely the experience we will probably be spending time

48:32later.

48:33Can I keep this side panel?

48:34that there we uh yeah there we go I can keep the side panel out by clicking on that logo uh and then it stays out it doesn't have to keep closing and I can just switch

48:42between these.

48:42Um

48:43develop I won't spend much time in.

48:45We didn't spend much time in

48:46the training, but it is essentially largely where you come to understand how to work with the embedding experience.

48:52and some of the more complex components behind the scenes.

48:54A large percentage of ThoughtSpot's customers, if not most, do engage the embedding capability, uh both as an internal asset and an external asset, so you can have it as an internal product size

49:05capability.

49:06Um

49:07you can even see some of these live.

49:08Let's go ahead and look at

49:09what some of these uh solutions look like.

49:11Um

49:11spotter code is obviously the

49:13AI capability allow you to

49:15sort of build these experiences.

49:16So this is like this is supposed to be an embedding demo.

49:19You kind of go to this little window and you don't really know that this is ThoughtSpot.

49:23It says powered by ThoughtSpot

49:24at the bottom, I think in this example, because it is um

49:27like

49:27just a test environment.

49:29But you can have these sort of different experiences, including the AI chat uh experience that we saw in SPOTA

49:37as something that you offer to your customers so they can come here and ask questions.

49:40And it's a really good uh

49:42capability.

49:43Like

49:43that spotter experience

49:45to me was actually, you know, pretty

49:46pretty sound.

49:47Um, so yeah.

49:49Uh

49:49there is

49:50there's a lot

49:51so much to cover.

49:52Um

49:53I'm looking at uh the account details.

49:56I don't think we need to do anything in here.

49:58So

49:58really as an analyst, the two places you're going to spend time are the Insights tab

50:02and the

50:03data workspace.

50:04To me, the thing that makes

50:07analytics platform

50:08at least in

50:10the next five years is

50:12going to be this

50:13window, the data workspace.

50:15How you bring data into the ecosystem.

50:17ThoughtSpot

50:18does support the ability to sit on top of things like semantic views in other systems, but you can also build pretty profound capabilities in the system itself.

50:26So um you can see here we've got uh DBT, um

50:30you can also bring in other

50:32Capabilities, but yeah, when we make videos about ThoughtSpot

50:36in the future, I'm gonna be spending a lot of time here in the data workspace just walking through each of the experiences.

50:42So connecting to data.

50:44uh

50:44bringing in uh

50:45tables, um

50:46configuring those tables,

50:48handling metadata, uh

50:50building data models that we can then use.

50:52Maybe we'll bring in bookstore with with Kirk and we'll try and build it out

50:56And then

50:57at the end of all that, when people

50:59come to ask questions in the agent or wherever they are, they get very, very good answers

51:05from the back of the

51:06that okay so yeah we've been going for 56 minutes

51:09is a little bit longer um

51:11a little bit longer because I

51:12spent a bit more time with ThoughtSpot

51:14so um yeah

51:14hope you found this useful

51:16Yeah, I'm gonna keep going.

51:17Uh

51:18tomorrow, I don't know what I'm gonna cover.

51:20Um, I think I said it was going to be Hex

51:23or Omni, one of the two.

51:25Um obviously already covered Tableau

51:27pretty well, so that's not in the list.

51:29But yeah, Sigma

51:30was yesterday, today is ThoughtSpot, Hex

51:32or Omni

51:32tomorrow.

51:33Yeah, and I'll catch you in the next one.