S5 E3: Byte: Technologies that pair well ish with Tableau

0:01Ravi, we're back once more.

0:03We didn't fail our obligation to stop for another what nine months.

0:08Nine months, yeah.

0:09The gaps are getting shorter and that's what we're in.

0:12They are getting shorter

0:13The the good thing about this gap is we actually uh pulled through.

0:16Um you were traveling, right?

0:17So we we c we actually couldn't record.

0:19You're in uh in a land far away where internet is banned.

0:24The internet's banned and everything is watched.

0:27Um no, it's it's it's always fun to visit these places because you realize I guess some of the pros and the cons.

0:35Um

0:36So yeah, that that was it was good fun.

0:38Um but yeah we mean when we missed our first recording date.

0:42Yeah.

0:42Um and then yeah, we've been trying to find whatever since I think

0:46We we we were messaging about this episode on the fifth of June.

0:50A month ago, exactly.

0:52Literally a month ago, yeah.

0:53We were supposed to record this a month ago, right before you went abroad.

0:57That was the deadline.

0:58And then

0:58I became ill and uh sorry like kids like just life happened and we were pushing it like too close and then I just said uh no let's let's do it when you get back and then

1:09I didn't know.

1:10And then football didn't come home.

1:11Yeah, football didn't come home, yeah.

1:13Are you wearing a football shirt?

1:15No, you're not wearing football.

1:15Ice Puma.

1:16It's a puma.

1:17I thought it might have been a football shirt.

1:19Yeah, football did not come home for all the heroes.

1:23It came home in 2020.

1:24It came home in 2021 when the women won.

1:26So um do you know who won Copper America?

1:29I think that's going on at the same time, right?

1:30Argentina.

1:31There you go.

1:32Uh was Messi playing?

1:33Or was he?

1:34Messi was injured in the first half

1:36So he came off early.

1:38Coincidence.

1:38Okay.

1:39Right.

1:39Yeah.

1:39And then I think he's done for international football, right?

1:42Oh yeah, I saw a picture of his ankle.

1:45Did not look good.

1:47But fortunately this isn't a football podcast.

1:50Good.

1:50It is not.

1:51No.

1:51The rest is not football.

1:53Uh thank God.

1:55So yeah, no, we thought we'd talk today about technologies that people use alongside Tableau.

2:01Obviously Tableau is our bread and butter.

2:03It's our it's our home turf, as it were, to carry the football analogy on.

2:07But we're taking some trips to the away teams and we're going to talk about things that you might see paired with Tableau.

2:13I think I will preface this by saying, look, we are not uh experts at these technologies, so

2:19We kind of wanted to have this open discussion as an initial gambit um to just talk between us, understand what we perceive uh these technologies to be, but obviously we're going to touch a bit on what they actually do.

2:31But we'd like to invite you, the audience, to um essentially get involved.

2:35Let us know in the comments.

2:36I love my comment section.

2:37People are so passionate.

2:38I the the one thing I love about the YouTube channel.

2:42is the people who are so polite, like this fills me with so much joy.

2:46They they open a sentence with like Tim, I absolutely love this video.

2:51But and then they put a timestamp like the

2:53like Stanisper was that Stanisper a few seconds and I'm all I'm always like man you watch till the end man you you got my heart

3:01And then they cut into me.

3:03They're like, you got this wrong and this wrong and this wrong.

3:06And in the end, they're like, but I really love what you're doing, man.

3:09Thank you so much.

3:10Keep doing what you're doing.

3:12And I love that.

3:13I love that.

3:13It is that stuff that keeps me going.

3:15So look, if that is you, um please engage with us on this podcast.

3:20We are going to try and keep to a time limit because we've got a long list here and I know Ravi and I are fantastic at just talking endlessly.

3:29So Which is why we started the podcast in the first place, right?

3:33Exactly, exactly.

3:34So what we're gonna try and do, I'm gonna I've got my iPhone here

3:36I'm just gonna kind of pull up the uh timer app on on my um yeah on my phone and we're gonna set a timer I think four minutes, maybe five minutes is more.

3:46Um

3:48We can be flexible with that.

3:52That's there's that Stannis Baratheon quote of anything

3:55before the word but is meaningless.

3:58Um so so the shit sandwiches you're getting sent to you um or on your on your comments.

4:04But no genuinely like if you if um old tricks Adam or

4:10Dbt Dave is out there ready to start our YouTube channel to rival Tyrant.

4:18Yeah, exactly.

4:19I I a colleague of mine at endpoint did come up with uh a databrick mason.

4:24I was like, that's very good.

4:25That's very good.

4:26Data brick mason.

4:28They should they should definitely get on with that.

4:30Um but

4:31Talking of Databricks, that is going to be our first one.

4:33So let me set a timer here, five minutes.

4:36Right, Databricks.

4:38What is Databricks?

4:41I I kind of describe Databricks as a data platform rather than a database.

4:47Would that be fair?

4:48Yeah, but it can do databasey things.

4:51I think this is where people get struggle with, right?

4:54You can have Delta Lakes and

4:56um lake houses within your data black Databricks platform.

5:01But I think it's the the biggest benefit is

5:04You can do a lot of things with it and probably the most important is it's built on Spark.

5:09So Apache Spark is just allows for mass parallelisation.

5:14Yeah.

5:15Yeah.

5:15That gives you speed.

5:17It's kind of like I always think of it as they they started their whole platform

5:23at the sweet time when AI and machine learning and data science was actually sort of coming along.

5:28We'll come to Snowflake later.

5:29Snowflake started too early to really capitalize on

5:33that sort of core core core structure.

5:35Maybe that's an opinion by the way.

5:36It's not a good idea.

5:38I think that that's what Databricks wrote.

5:40They wrote that is a big thing, yeah.

5:42Or area of like being able to actually apply and give you examples of

5:47Where it can solve your problem.

5:49Okay, this is a problem that you have.

5:50This is how this solves it, and this is how it gets around it.

5:53Um

5:55And you know that there's there's actually quite a common meme going around the social media platforms right now of like, what does Salesforce do?

6:01Like, can someone explain what Salesforce is and what it does?

6:05Yeah.

6:06And this it's almost what you felt with

6:08With a platform like Snowflake, we'll come back to Snowflake, obviously.

6:11Yeah.

6:11But these platforms, and I think as you see as you said, the timing of Databricks to come almost emerge onto the scene and say,

6:18We're a platform, but more importantly, we're a de-aggregated platform.

6:22You can host us anywhere.

6:24We can connect into anything and we can connect out into anything.

6:28We're just the people in the middle that allow you to do things quite cleanly

6:32Uh and they kind of realize the role of the warehouse in that, right?

6:37They kind of they kind of I think what's the word?

6:41They reimagined the warehouse in this in this very slightly different way to Snowflake.

6:47Same sort of direction, slightly different execution.

6:50leads to much, much sort of deeper integration.

6:53And I think it's definitely got sort of the hearts and minds of um a lot of companies because of that slightly more integrated

7:00more sort of disaggregated sort of nature that you talk about.

7:03So um super interesting.

7:06Have you used the Ravi?

7:08Yeah, yeah, I've used it.

7:09It's um I think the nice thing is you you can write notebooks where you start off in Python, do a bit of SQL, tack on a bit of R, and then execute it all as part of a

7:20wider warehouse query um like a datafectory query or you can call on the notebook via a API.

7:27So would would you say the learning curve is steep because of all those prerequisites you talked about are Python

7:33If you've not really sort of dabbled in, let's say, I'm gonna say data science generally, but also if you've not been a multidisciplinary, I can't say that multidisciplinary

7:48That's the one.

7:48My multidisciplinarian in your approach to learning analytics, it could be quite a steep learning curve because you've got a lot of, I would say

7:57front facing things to learn and principles as well.

8:00It's not just each of those things, it's how do they all come together, right?

8:03I'd say that that's me, right?

8:05Like I I don't I don't see myself as a data scientist and I don't I I didn't understand how a

8:10I can't do a V lookup or an X lookup and I don't understand how pivot tables work until Sai.

8:15AI is coming for your job, mate.

8:17Exactly.

8:18Um but I didn't really understand it.

8:19And then I think being able to explain in

8:23You know, the the fact that this thing does this, this thing does this, this thing does this is quite clear.

8:28But yeah, I I think I I'd agree with you.

8:30Um there's a lot that you can do with a platform like Databricks.

8:34They don't really get into.

8:36Like you can just use surface level.

8:38Here's my data lake.

8:40I want to query it, and I'm going to get out.

8:42Yeah.

8:42But you can go really deep into it as well with orchestration APIs.

8:46It's it's a technology that I I feel like I I need to find a good use case for in the future.

8:51I don't feel like I've

8:53It's not for a hobbyist, right?

8:54Like this is purely enterprise, scalable, things that are just have to be constantly running and updated.

9:01There's no tableau public for Databricks here, like.

9:03No.

9:04That's called just it's just Python or Jupyter.

9:07Yeah, exactly, exactly.

9:08So I I think that's um a pretty good touch on the data bricks.

9:12I think if we were to talk more about its integration with Tableau, they announce a new connector, Delta Connectors.

9:18Um

9:19literally live as of a couple of days ago.

9:21So that's a really good way of connecting to um to Databricks.

9:25I think the vantage trend is that it it it it it should make that easier because previously it was quite hard, right?

9:31Yeah, it it makes it easier for sure.

9:33And like it it's an exciting capability that you can tap into.

9:36But in order to set up you need a bunch of

9:42Um in order to add it up you need a bunch of like the relevant information and keys, so you might need to actually reach out to your uh administrator, which if you're a

9:51you know, Databricks user within a large enterprise environment, you might not know who that is.

9:57Um so this is where you get end up in that weird

10:00uh moment where you know tableau used to live of being shadow bi um and maybe you end up with shadow IT to try and make your Delta Lake connect to work.

10:09The the other one that I'm I was kindly reminded of

10:13uh when speaking to him at the tableau conference uh by Thomas and Hahn was uh the table extension which we talked about really early on.

10:24We've got a previous pod on this actually, but we'll come back to that, yeah.

10:27Yeah, and and this can link into and call on Databricks models.

10:32So you you have a model in Databricks, you can call on it and extend your data tables.

10:36Yeah.

10:36Um and again like that is a super powerful the iceberg for Databricks is big.

10:41Yeah.

10:42I I think that's sort of one of the challenges with Databricks potentially.

10:47It breaks away from the tableau mold and might be slightly tougher because it does it does require you to look beyond the

10:55the drag and drop, like the the the sort of utopia that Tableau presents you, this this is not you think.

11:03Yeah.

11:03You have to get deep into that.

11:05This is not for the like drag and drop sort of enthusiasm.

11:08It's not for the

11:09you know, window calc dabbler, this is uh this is like a hardcore tool, a little bit, yeah.

11:15Um but when when you if if you've got a great data engineering team,

11:19They're probably using this, yeah.

11:21They will be able to like understand your query and then help you translate it so you just deal with very pretty

11:28Yeah.

11:29Delta tables that you just connect to and refresh as you normally would.

11:32And we we then end up living in a rosy environment.

11:37Right, the timer went off a minute ago.

11:39We went over, so we clearly have a passion for data breaks.

11:42Let me reset the timer.

11:43Start it again.

11:44Let's go to Snowflake, because we brought it up again as we were talking about that.

11:48Um snowflake, I feel, was right at the curve.

11:51But I just want to call out, you pointed out Snowflake to me.

11:54I don't know, I can't literally at this point.

11:57Like TC17.

11:59TC17 at the tobacco dock.

12:02You actually took me to their stand and we watched the demo and I I stood there and I was like

12:08I just don't get it.

12:09What is the like I don't get it?

12:11And you explain to me it really, really well.

12:14And what it what it what it shows to me.

12:17I just wasn't far along in my learning journey to really appreciate it for what it was at the time.

12:23So that that was the second time.

12:24I think we I I took you to their stool twice.

12:26The first time I was like, I've I've heard this thing and I think it does this.

12:30Let's go ask the guy.

12:32And I think that we were at TC in Vegas and we went there again.

12:36I think we both left that being like I still don't know why I should use it over a normal database.

12:42Yeah.

12:42Right.

12:43I get it.

12:43Yeah.

12:44But I just don't I think there's a really small niche of people that this is really useful for.

12:49And that's just people that don't

12:51Yeah, th there's just a smaller niche at the time.

12:54Um But I'll also say this, since then they've massively built on their platform.

13:00in a way that, you know, having also been the one who's like, I don't get it, I now have a course on LinkedIn introducing people to it.

13:07So they've massively built on the platform, massively in in a way that I think

13:12They do come in competition with Databricks for a lot of what I would say the same type of workloads with data, not specifically database workloads necessarily, but the same types of work, the same types of computations.

13:26Um, but they have built they really like the way I think of Databricks is that Databricks have built um a platform, you deploy it how you want, where you want.

13:36Snowflake.

13:38They built the software, they're running it on a platform, AWS, and you deploy it for you.

13:44It comes in multi-flavors, yeah.

13:47But it has to be on a web platform.

13:49So it's very much like a SAS version.

13:51I know this is a bad description, don't don't come after me.

13:54But it's very much a SAS version of what Databricks is trying to do, right?

13:57Like

13:58it it's kind of like a lair that abstracts the hard work of doing that, but really people just want to get like um uh close to the to the metal as it were so anyway um

14:07How quickly can I find this book in my library when I want it and access it, read it quickly and then put it back?

14:15And just take this out, do stuff, put it back.

14:18And then it's like at point of

14:21query or point of access, you're only getting that doing something with it and then you're it could then disappears.

14:27Like I don't know if you watch um Rick and Morty, but there's a thing called a Me Seeks box.

14:33And then you press you press a button, a little thing appears, you then say, I want to do this.

14:38It does its thing and then it disappears.

14:40And that's exactly what it does.

14:41That's all it does.

14:42Like Siri, the data.

14:45As my toy.

14:46Or an auto or one of those auto AI like auto GPT?

14:51Auto GPT.

14:52Where it like spins up micro anyway, sorry, we're going off on tangent.

14:55You triggered my uh Siri as well.

14:57and I've just triggered it again.

14:58Jesus, I need to there needs to be a way of disabling Syria for a period of time.

15:02This is um this is ridiculous.

15:04I'll stop saying the word so um

15:08It is interesting, I think, that um hold on.

15:11Let me just stop a second because this thing is just still recording what I'm saying.

15:14This technology man!

15:16Oh, I just hope Apple Intelligence saves it, because that is ridiculous, but okay.

15:21Um back to the back to Snowflake.

15:23Thanks, Siri, for that interruption.

15:26Back to Snowflake.

15:27I think the other thing they've done is they've massively leaned into

15:32B2B workflows as well, right?

15:34Like the ability the the data clean rooms, the ability to share, doing that at scale.

15:38I'm definitely I like I've come across instances of industries centralizing on Snowflake, which is

15:44really good because you kind of get the economies of scale of go of doing all of that stuff.

15:49And then the more recent partnerships over the last couple of years with NVIDIA.

15:54to start doing containers and compute clusters in the cloud as well, trying to really lean into similar kind of spaces, data bricks coming at it from a different angle.

16:03Really powerful.

16:04I is there

16:07I always wonder with Snipe if I always wonder if they were too early, just a tad too early, if that makes sense, right?

16:15Like

16:16I I you know, the world we're living in today and uh and and I can't really say this because if if you sort of tweak it a year or two here, then actually just at the right time.

16:30But

16:31I I definitely didn't come across Databricks at the same time I came across Snowflake.

16:35But the rise to um oh, there's our timer for Snowflake.

16:38The rise to like let's say

16:42Being well known for Databricks seems to be in shorter.

16:45Is that is that fair to say?

16:46Like I feel like Databricks has had like a five-year run-up where a snowflake really did or a great marketing team.

16:53Okay, fair, fair, fair

16:55I saw Shade on the Snowflake team, but yeah.

16:57No, yeah.

16:58I think um the thing is when Snowflake came about, their competitor was Exosol.

17:03Like Exasol was one of their competitors.

17:06Yeah, exactly.

17:07Exosol, right?

17:08Yeah, where did they?

17:09So you had Exosol in there, you had Metrika, Metrika, HP, the HP

17:15One and then you still had like Hadoop flying around.

17:18Vertical not Metrica.

17:20Metrica's a football company.

17:25Um fair, fair.

17:26Yeah, so uh an HV vertica, SAP and then Excessor.

17:31Like so you're coming into a space where everything is actually about speed of query rather than like

17:39storage versus compute, which is what they're doing.

17:42But everyone just focused on the like, you know, classic, you know, gigahertz, megahertz race and

17:48suddenly Databricks comes along and say, actually that doesn't matter as much.

17:52Well it does, but really all these other people.

17:54If you just take out the process part, the transform and extract bit, and then just say, but you just need this, right?

18:00Go get that then and let us do this.

18:03Yeah.

18:04Pretty deep integration for yeah.

18:07Pretty deep integration for Tableau.

18:09Um and I I think Tableau use Snowflake at on, right?

18:12Oh really for their for their own um

18:15marketing data and uh uh clusters.

18:18I keep hearing product developers talk about like uh obviously they keep their data in lots of different places, but Snowflake is there like big um

18:26the big big thing.

18:27I they've talked about it at conferences, so I'm not I'm not hopefully I'm not saying something I shouldn't be saying.

18:32Uh I'm referencing conferences just as like a standard point 'cause I like if I could find it in a conference talk, which I can, then it's public knowledge.

18:41Um but anyway, um super interesting.

18:45I I don't think I've seen as many

18:49use cases with Snowflake as I thought I would have seen if that makes sense.

18:54And by that I mean you know how with um Databricks you've got this, you know, you talked about table extensions.

19:01With Snowflake, I don't know if it's just because it's mostly within enterprise or mostly within certain sectors like fast moving consumer goods, but all these, you know, data clean rooms, all of this stuff, I haven't seen that in the tableau world as much.

19:13I don't know if it's just because I've

19:14I'm not that deeply embedded with the like the pro use case for this data masking as an example.

19:19Like that's a really good idea of for something like

19:23You know, with Tableau where we come across these sort of common privacy questions all the time, imagine being able to just do like a parameterized um masking operation, right?

19:33right inside of the query so that you're just connecting live and depending on who the user is with user um user filters or uh forget the name of the role um

19:43Not roles, I forget the name of the feature, but anyway, user user attributes, that's the feature.

19:48So with user attributes being able to dynamically mask certain bits of information because you're able to look at the user attributes and just modify the query.

19:55All that stuff

19:57Luka used to be able to do this.

19:59Look at LookML.

20:01Yeah.

20:01I think LucML allowed you to parameterize and do Dust

20:05Like which is which is hilarious, right?

20:06They sell them says as a visualization tool, but the thing that you everyone always talks to you about is LookML and the fact that you can do loads of like parameterized

20:16like data filtering anyway.

20:18Um I think the the reason for that I would possibly argue again having less knowledge of Snowflake than you is probably that deaggregation point.

20:28Like

20:29If you think about platforms as aggregators, right?

20:32Like you had the period between 2014 to 2019 where you're like, we can do everything.

20:37We are a platform, just one stop shop.

20:40And Snowflake almost became that, but then Databricks were like, yeah, no, we're like a Christmas cracker.

20:45You can open us at both ends and we'll just stay in the middle.

20:48You want us over here, we'll go over there.

20:49If you want to go over here, we'll go over there.

20:52Um whereas I don't know if Snowflake can do that, therefore extensibility in the same way we talked about with table extensions, for example, might not be the case.

21:00However, it might.

21:02Snowflake Sleep Sneave Sneeve?

21:04Snowflake Steve get in touch.

21:08Let's have a chat with you on the podcast.

21:10Absolutely, absolutely.

21:13Right, let's um let's take a detour.

21:15Let's go to all tricks very briefly.

21:18Very briefly.

21:20Well

21:21What is there to say?

21:22Um I'll t I'll I'll start.

21:24I'll start because I use Autrix every day, so um I'll I think if you let me re let rip, I'll I'll go for too long, but yeah.

21:33I've not used Alteryx for three years now.

21:36Right.

21:37Um possibly longer.

21:39Um I

21:43But back in 2016, 2017, if someone I think someone did ask me, like if you've got to pick one of these two to back to go long.

21:56Is it Ultrayx or Tableau?

21:57I would have gone Ultrix because there's a lot more use cases and it's a lot less crowded field and it's really easy to use.

22:04But the price point, the limitation over the engine, and

22:09The fact that code like almost AI would have killed it, like destroyed it completely, right?

22:17Because suddenly no code isn't that interesting, or drag and drop isn't as interesting.

22:21Yeah.

22:21Because you can find

22:22You can have AI assistant supporting you on your code.

22:25Databricks AI assistant jumps at you when you know you need it, but it does solve your problem, for example.

22:31Yeah.

22:32But it's really fast and I think the in memory for someone doing stuff locally for themselves

22:39To do, let me just do this, this, this, and this, and just as I'm thinking, I can drag something and drop it, get the answer.

22:45Okay, this is the schema I want.

22:47Now I'm going to show this to my data engineer.

22:49Can I build this?

22:50Go productionise it.

22:52But for that fast iteration, incredible.

22:54For the ability to implement code and automate and do batched macro jobs, fantastic.

23:01Systemic jobs, amazing.

23:03Like if you want to generate and burst PDFs.

23:06and run some code and then bring it all back together.

23:09Yeah.

23:10Yes, we can do this.

23:11Spatial.

23:12What it he lives, breathes, and eats.

23:15Um but it but it just didn't evolve.

23:18Um ultrayx.

23:19next, ultrax server, oldrix cloud, whatever you want to call it, was just never the thing you wanted it to be.

23:28Yeah, like I'll I'll add to what you said at the beginning.

23:32Yeah.

23:32I'll add to what you said at the beginning, which is, you know, if you chose something 2016 to go long, it would have been all tricks.

23:40To this day, if I only had one tool I could use as a data analyst, like one tool, I had to pick one tool.

23:49I couldn't open any other data product.

23:53still be all tricks because I can muscle my way through any problem in that talk.

24:00And they're like

24:01Maybe because I've used it for so long, I've just have I've I have I'm like I've got it I'm in tuned to how you're not wrong because like data densification kills you in some parts of Tableau.

24:10There's so many times you hit the wall of you need to

24:13pivot your data and you cannot do this unless you pivot your data.

24:17The other thing is in altrix you just you just pivot it and carry on.

24:23We'll come on to this in a second because DBT is on the list, right?

24:26But I'll say this.

24:28I have never met a tool that allows for accelerated ability to solve last mile problems, right?

24:40Like you've you're you're you've got the data, you've got it so close, but there's you you just come across a use case, some something in all the data is right in front of you.

24:49It's just not in the right format, it's not in the right shape, it's not in the right

24:54It's not in the right place to do the thing you need to do.

24:57Yeah.

24:57You don't want to go back to the data team.

24:59It's still all tricks.

25:00And DBT and all these other tools don't solve that because what they do is they put up a new hurdle.

25:05A new language, a new way of processing.

25:06Whereas Ultrax allows you to stay in that sort of visual frame of the data, like rows and columns, right?

25:12And you can just get the data to where you need it to.

25:15Still to this day

25:19the best tool at that.

25:20I know there's other players, prep, we'll talk about Ensos as well later on, but it's just it's just such a loss of this is this is such classic like

25:29You know when you're just like ah but it's so good and I love it so much but it's not the one and you know it's it's hard to go that it's that is my heart sort of.

25:44It's so true, it's so true.

25:45And we said it wouldn't talk long about this, and here I am basically writing a love letter to Autrix.

25:50Um but nonetheless, listen, they've not kept up with the cloud.

25:54They missed their AWS.

25:57Pricing structure killed them.

25:58Pricing structure still kills them today.

26:00Still kills them today.

26:01Um and there there is

26:07There's just a lot of very um I think difficult difficult things for people to get if they've not used a tool before.

26:14Like if you're sort of introducing some new tool tricks.

26:17Uh that philosophy in Altrix is quite hard to understand.

26:21It's sort of derived from years ago, but in today's world it just it just really grates against that way people think about data.

26:27But I still stand by it.

26:29If all they did, I know it sounds simple to say, if all they did was they took their core product, Dortrich Designer.

26:36Push it into this modern world.

26:38Put it in a browser for goodness sake.

26:39Put it on a Mac, for goodness sake.

26:41Like bare minimum, put it on a Mac.

26:45Bring it to data scientists in a fresh way, right?

26:48Bring it to data analysts in a fresh way.

26:52Make it play nice alongside the things they already do.

26:54Tableau, Power BI.

26:56Just make it a companion to those things.

26:58It doesn't have to be a product of its own.

27:01I've got a killer there, right?

27:03Imagine, imagine the simplicity of

27:06old tricks and then once you're done you can output it into a jupyter notebook and it will do everything for you.

27:13Yes.

27:13You can then push it into R, it's done it for you.

27:17You want it in Python?

27:17Here you go.

27:18You want to do this all in SQL?

27:20This is the the things you can do.

27:22They have it in in databases literally that you do your data prep, you do your thing, and then there's a little thing you're saying, oh would you like me to write the sequel for you?

27:29Donna.

27:30It's literally there.

27:32Like just

27:36Lots of people describe it as a Swiss army knife.

27:40Just lean into that man.

27:42And build a platform, the cloud platform that's deserving of designer, rather than trying to make the platform its own separate product.

27:51Anyway, we need to.

27:52Still sponsoring McLaren, are they?

27:53Are they still sponsoring McLaren?

27:55Yes.

27:55Absolutely.

27:56A great sponsor.

27:57Keep seeing Lando Norris.

27:59Well the Aldrichs are getting great value for money because they bought that when McLaren was nowhere.

28:03And I think I believe that

28:05Not even in the midfield.

28:06Yeah, exactly.

28:09Probably the first two years of that sponsorship, that was probably looking sketchy and then

28:13in the last year it's come through real good.

28:15Like the amount of times Rando Norris has been on on the podium is unbelievable.

28:18And I keep looking at the jacket.

28:20The sponsorship position moves.

28:21The sponsorship position moves.

28:23It did used to be like quite prominent.

28:25And then it's now on the sides, like in the long list of names.

28:28Sometimes it's on the front nose of the car.

28:30So it's not got a permanent spot, but it does move around.

28:33So

28:34You know, I'm still waiting for that race where it's like right on the front nose and like it's a race.

28:40Or it's a piece of debris that's flown off into the distance and you just see all tricks fly across the screen.

28:46Uh that is a that is a slowbo moment I'm waiting for.

28:49Something to happen and then just see the old tricks like in that it's just like brilliant.

28:56Um right, let's let's go on to DBT.

28:59DBT's next DBT is next.

29:01Um

29:03Gosh, I think DBT started off with an open source tool.

29:06For anyone who's wondering what is DBT?

29:08I didn't even describe Altrix.

29:09Um I'm really sorry.

29:10We won't go back to that.

29:11Drag and drop tool makes it look at it.

29:13Drag and drop visual analysis visual data cleansing breath.

29:16wizardry predictive wizardry mapping you kind of did actually so let's just carry on dbt um I kinda the way I describe dbt is imagine SQL

29:28But with smarts on top, basically.

29:30That's that's basically how I describe DBT.

29:33And the big problem they have is I think they built a great open source product.

29:36Everyone was like, this is amazing.

29:38And then they tried to build a company on top of that.

29:41And everything went to their cloud platform.

29:43And I think that's been like a difficult transition for people to understand, right?

29:48Um but there's a lot of good principles.

29:50Like you know, there's you know, I think if anyone ever talks to you about semantic layer, they're really ultimately talking about some of the capabilities in DVT.

29:58I know other talk do the same.

29:59same thing.

30:00But nonetheless, yeah, they're talking about DBT.

30:02So I mean do you use it in your work, like in in what you do?

30:06No?

30:06Interesting.

30:07It's it's it's it's another one of those things that you hear the name and then you ask your trusted circle of

30:13Data nodes being like, should I care about DBT?

30:17And then they tell you, yeah, or no, or I don't know.

30:20I keep hearing it as well.

30:21Tell me when you find out more.

30:23Tell me when you find out when it's good.

30:24I've used it for a project.

30:26I can't talk about the project, but what I I I came into this project midway through.

30:32So I like if it was my choice I wouldn't have built this on DBT because I would have built it in Altrex.

30:37And then I would have struggled with Ultrix.

30:39Of course, I do everything in Ultix as a theme here.

30:42But what was interesting about experiencing DVT, and I experienced two flavors of DBT, DBT Core, which is the open source.

30:50version and dbt cloud which is the paid version runs in the cloud my experience of cloud was fantastic really good like the lineage was oh chef's kiss I was like

31:01Damn, this is nice.

31:03Like, where is this in Nortrix?

31:05Like I know they had a data product.

31:06Or any other data product to have that clear version control C I C D.

31:11Yes.

31:11Are you this this this model here, this Python thing you've got

31:15Yeah.

31:15It goes here and then goes here.

31:17This is take a step back.

31:18Here's everything.

31:19Exactly.

31:19Yeah.

31:20Like beautiful.

31:22And then the thing I said earlier on, if you recall, last man analytics.

31:27killed me.

31:28Like the amount of times I got into the workbook and I was like, God damn it.

31:33I need to go back into my model and like

31:37Fiddle with this and fiddle with that.

31:39That's my wife calling me.

31:40It's fine.

31:41Fiddle with this, fiddle with that, and then I'm like, oh god, I did to build the whole model.

31:45Like Jesus, I could have just written a nice, beautiful, like

31:48LOD right here and just solve this.

31:50But no, we're doing things the DBT way.

31:55A product developer we both know well, who happens to work on calculations, described it as

32:01uh uh communism for data sources which is which I which when he said it I didn't get it but like when I think about it I'm like actually

32:11Nailed it.

32:12Absolutely nailed it.

32:13So so true.

32:15I'm just looking at the docs now.

32:17Um and any docs that automatically goes into dark mode, you know the people that are using it.

32:26Fine, they've thought about it.

32:28They've thought about it.

32:28You know what?

32:29Like the people who are gonna be deep in these docks want dark mode immediately.

32:33They don't want to start off with none of this white white stuff.

32:37But no, I think I think the from the website and from what I've heard and the the reason I sort of almost asked was exactly that, the back the fact that you can deploy at you can deploy many things.

32:49And then follow it almost like this meta layer.

32:52Um so you can do the nitty-gritty, but you also get the meta layer.

32:55But you're not nimble because it's again designed for enterprise, not for your mom and pop doing your Strava analytics

33:03Like you're not with Strava analytics in here, right?

33:06Same same with data, for example.

33:07But Snowflake, you might, right?

33:10You might use a community version of

33:12Yeah.

33:13Snowflake to host your your personal data.

33:15Yeah, you can just start up a new trial every every month with a new email.

33:19That's how that's how you get free Snowflake.

33:22Um

33:23But if I go back to DBT, I think the bit I really struggle with is DBT Core.

33:29And man, DBT Core, that is painful because you don't get all the same niceties.

33:36Um

33:37It's a bit it's a bit like that sort of rough experience you get with um with um No Explorer, no match, no semantic layer.

33:49DVT um

33:51I think it's still quite promising.

33:53I this is not gonna sound nasty.

33:56Is it a product or is it a feature

34:00Jesus Christ.

34:02So it sounds like you're saying DBT is going to get bought out by Databricks or Snowflake or someone and absorbed into their product.

34:10It is Or the principles.

34:12Or or or do you think that the principles shouldn't it be that?

34:16Should surely the way DBT works is exactly that.

34:19Tell me that it doesn't belong inside of an analytics platform

34:24that's decoupled, that it is agnostic to you know where it's come from.

34:27It's the thing you put on top of your platform to orchestrate it.

34:30It doesn't matter what it is.

34:32It's a feature, surely

34:35Yeah.

34:35Trying to think of I'm trying to think of another product that that that almost does that.

34:40And it's usually freemium stuff, right?

34:42Like it's like oh yeah yeah.

34:43Correct.

34:44And maybe it's explains why people struggle with using it, 'cause you just find well, I just do this myself with my own little stored procedures in SQL and

34:51Well this stuff, right?

34:52Like so yeah, m a brutal ending there.

34:55Is it a product or is it a feature?

34:57We'll come back to that later.

34:58Um

34:59Enso.

35:00Enso makes anything.

35:03The new kid on the block.

35:04Yes.

35:05Interestingly, it allows me to go back to my love, Altrix.

35:09Because it's got many of the same technical team, right?

35:12The very very heart of the product.

35:13Adam Riley, Steve Harding, I think, as well.

35:16Yeah.

35:17Steve Harding.

35:18The original creator of the Ultra Channel.

35:20James Dunkelly.

35:21Yeah.

35:21James, exactly.

35:22Exactly.

35:23This is like the greatest hits of let me even go further and say, this is that era

35:34Ned Harding, sorry, Ned Harding.

35:36No, I go back to like 20 is it 2015, Ravi?

35:39You've got Libby, you've got Ned.

35:41You've got um uh Adam Riley, crew macros.

35:46There was just a sweet spot there where the Orfrics was like, use a British term.

35:51I've got a greater side here.

35:53So um Go for it, go for it

35:55Yeah.

35:55I think it was it was it was at one of the Alter X conferences in London.

35:58Uh and Will Griffith's good friend of mine, former data school consultant.

36:02Yeah.

36:03Um

36:04I think it was him or Ben Moss created a ma Ultrix macro called The Scrambler.

36:10Because the YXMD file, the Ultrix, um

36:16file that you use to open up workflows is just XML.

36:20So obviously if you can find out the X and Y coordinates of everything on your canvas.

36:25and then just randomize it, what you'll get at the end of it is just all of your tools everywhere when you just put it through the randomizer, the scrambler.

36:33But it still works.

36:34But it still works.

36:35You can still run it.

36:36It's just in a horrible order.

36:38So Ned's Ned's in London, and we're we're talking to him, Adam Riley, James Dunkley, and Ben's like, show Ned the scrambler

36:48Um sorry go through it, show him, explain it to him.

36:52And Ned's words were, when I when I you know wrote Bill designed Alteryx back in the 90s.

36:59Didn't see this as one of the use cases.

37:07Just love it.

37:09Love it.

37:09Love it.

37:10So yeah, and so um cloud native, uh I think it brings, dare I say it, the philosophy of altrix, but to a modern to a modern

37:22To modern world.

37:24And I'm not going to go as far as saying it is all tricks.

37:26I'm not going to go as far as saying um, you know, you know, they're just forking altrix or whatever, because I don't think they are.

37:33I feel I've seen some posts that um uh

37:38Adam Riley, his name is so hard.

37:40I just think of him as Mr.

37:42Crew Macros.

37:43That's in my head.

37:43I just want to say the guy who built Crew Macros.

37:46Chaos Razor.

37:47Yeah.

37:48Um, I always forget the name.

37:49Adam Riley, Adam Riley, Adam Riley, I won't forget it now.

37:52Um every time I see Adam Riley do a demo on LinkedIn, there's always a little bit more thought that goes into some of the simpler tools.

37:59that we'd we we've never seen an Ultrex.

38:02And it goes all the way into UX and UI.

38:04And it actually reminds me, this is this is what the core problem is.

38:07You know, Ravi, you have not used Ultrex in five years.

38:09Let me let you into something.

38:11Not much has changed.

38:12You don't miss much.

38:15And it's every now and then the ultra is crumpri.

38:19And I'm like, that's the same

38:21It's literally the same interface.

38:23Again, you get to make the same argument Tableau.

38:25What's changing Tableau Desktop, Tim?

38:27What's changing Tableau Desktop?

38:28Everything.

38:29Jesus, don't get me started.

38:31Don't get me started.

38:32Don't get me started.

38:33Don't get me started

38:34It's not for the dashboard designers anymore, Tim.

38:39To the core.

38:40It's to the core.

38:43Oh no, no, I've been on a roller coaster ride about that.

38:46We'll save that for another podcast.

38:47But going back to Ense

38:50It just feels like some of those initial places.

38:52It's very clear the product team did want to go, did want to go and understand and discover and reimagine and re-envision.

38:59They're getting the opportunity to do that in ENSO.

39:02And I think, you know, if you'll listen to this, go check out ENSO and don't come at it from an altered perspective.

39:08Go with a sort of a neutral mind, a clean mind, as it were.

39:11Evaluate it for the product it is, and I think you'll find um it's quite surprisingly nice.

39:16Um so it's quite good.

39:17I I think again the the the benefits of Tableau and Alteryx have always been you can think and do in the same

39:26flow right like you you you you end up in a flow state and looks like enzo are almost taking this yeah um from a flow state just looking at the um the documentation

39:39In the getting started, there's really a CSV, passing selecting comments, how to use ENSO documentation.

39:45The fourth one is your journey from to ENSO brackets from Alterix.

39:50Oh yeah.

39:51They're going for it hard, right?

39:53And uh they're not the only ones.

39:55Um we'll touch on data IQ maybe in a bit.

39:58Um

39:58uh also going for the oltrex customer very very hard.

40:01I think everyone's just looking at Altrix and go, yeah, we'll have a bit of that.

40:05Um who is it?

40:06Um FiveTran.

40:07No, I kind of think feel like had their moment taking a bit of Altrix's uh lunch.

40:12Um

40:13I love the philosophy behind Five Tran, which is, oh, to hell with the freaking macros and the flays.

40:19We'll just tell you the answer.

40:21We'll just do the work for you and just give you beautiful tables.

40:24And by the way, you want DBT here.

40:25We've even built these DBT models for you.

40:27uh you know off you go like I am still shocked by how easy it was to go from not a five tran customer

40:37To a paying five strike customer who within 30 minutes had exported my entire YouTube analytics for the last six months to Snowflake in 30 minutes.

40:51Now I know speed or ease?

40:54Both.

40:54Was it just doing the work for you?

40:55Just both.

40:56Just both.

40:57I literally turned up to Snowflake 30 minutes later.

41:00I looked at the tables.

41:02I looked at the architecture, I looked at the models it done, and I was like, perfect.

41:06I I've like I've got what I need.

41:09I I don't even need to think about it.

41:10I just literally opened Tableau.

41:12And a problem that would have taken quite a lot of considerable effort and time was just solved.

41:17Now, was the cost okay?

41:19I know FiveTram is quite expensive.

41:21Airbite is cheaper, does the same thing, of course.

41:24But I also wonder like why didn't Altrix just lean into some of that as well in the past, right?

41:29Anyway, keep going back to Ultra.

41:31Keep going.

41:31Can you tell that I love the product, but I hate where it's gone?

41:38You can just play a violin in the background.

41:39I'm Tableau Tim and I'm out here crying my soul out for altrics.

41:43Okay

41:49You know this better than anyone, Ravi.

41:51It's probably because of the amount of time I've spent in that product solving stuff for customers, right?

41:56Like

41:57Those FMCG projects, Tim.

41:59I'll mind those are some numbers.

42:03Recently insurance

42:0590% of my time was in Ultix solving huge problems.

42:0910% of my time was in Tableau because I solved it all in all in Ultrayx.

42:14So anyway.

42:15We've moved on.

42:16Enso, check it out.

42:17DBT, check it out.

42:18Datriq, check them out as well.

42:21We've got the cloud platforms.

42:23So I'm going to rattle them off.

42:25There's only really three you need to be aware of as your AWS, Google Cloud Platform.

42:32Even GCP can almost like start to be like, yeah.

42:36Yeah, exactly.

42:37Because yeah, like I I don't know any I don't know many positive people when they use GCP.

42:44It's fine, it's hard to use, it's all cloud.

42:48But in the AI world

42:50Where you gonna run your sheep?

42:53Where are you gonna run your uh um um AI compute?

42:57Um

42:59God, I can't get my words out.

43:01What is the word?

43:02Where are you gonna run your transformer models?

43:05That's what I was looking for.

43:07Um so Google Cloud Platform, transformer models, this is where you run them

43:12I I think this this is about to change.

43:15If you talk about Azure, Amazon, GCP, Azure is the only place you can run Chat GPT models.

43:21Fine.

43:22Lots of people are going there.

43:24Um Google is the only other person.

43:26I think a lot of people want there to be other competitors to Google.

43:29So I think GCP might have some benefit in being the only other place.

43:35But Amazon is not really part of that, right?

43:37Amazon's not part of that sort of, let's say, dialogue other than it's still the place you go to run compute.

43:45to prepare your data to be able to run it through AR models.

43:49So they they are like a transitory platform for all of that stuff.

43:53But yeah.

43:55It's sort of a so so so my my view is that the reason I don't see much future in in GCP or like

44:07Um you know, uh sorry, not in large future.

44:10I think it'll be used, but it's just always useless.

44:13What I'm what I'm

44:15Why I think AWS and Azure sort of live longer is because they have bigger customers that use them for everything.

44:23And the scale of it, like in the coverage they have, is just massive.

44:27They're kind of baked in, aren't they?

44:28Yeah.

44:28They're baked into the case.

44:30And and like for example, in in order to have Azure any like anything with Azure

44:35Be it you just need the platform.

44:37And once you have the platform, you may as well get Microsoft SQL Server.

44:42You may as well start using Power Automate.

44:44You may as well start

44:46Going through all of the different things you have that are available for Microsoft.

44:52And AWS is the same.

44:54Like it's really easy and really it's a really nice platform to use.

44:58And it's like

44:59When you and me both did the certification around the same time, right?

45:02Yeah.

45:02Like it's it's such a nice platform to learn.

45:05It makes sense completely and you can you can build from there.

45:08And again, all three of these to an extent.

45:12work quite nicely with Tableau.

45:14I think GCP is the only one that doesn't because BigQuery is very difficult to connect up to Tableau, in my opinion.

45:24I've never had a positive experience with BigQuery in Tableau.

45:29But go Google Gregory, GCP Gregory will come on this podcast and talk to us.

45:34I said George.

45:34GCP George.

45:36There we go.

45:36GCP George will come on.

45:38Greg or George, one of the two.

45:40Yeah.

45:41I I don't know.

45:43I still

45:44I I think good GCP still lives on purely 'cause it's Google and Google probably run a ton of their own stuff off G C P as well.

45:50And like and then Apple run a ton of stuff off of GCP and Amazon to be fair.

45:55And then a lot of like what I would call

45:57uh Spotify run off GCP.

45:59And so if I start to list the companies that run off GCP, a lot of the companies that do that purely do that because they want to be able to have something against Amazon because Amazon

46:08like a big a big big share of the cloud so gcp lives because amazon exists right it's not it you know it will never die because you always need something that isn't as your

46:19Against Amazon.

46:22I mean again I I I feel like of of the of the three, if you could talk to talking about their database and connecting to their data.

46:29I have had no joy with BigQuery.

46:33I've had very little joy with Redshift and Athena.

46:38And SQL Server's been alright.

46:41Um, so who of the three, if you're connecting data, I'm I'm I'm I'm the I'm an Azure man.

46:47Fair.

46:48To start with.

46:49And and it's gonna be hard for you to move out of that, even if it was to Amazon or something like that, right?

46:54Because it's the enterprise.

46:55Yeah, yeah.

46:56So if we talk about Tableau and we talk about these technologies, right?

47:00Azure, well in Azure you've got Microsoft SQL Server, you've got obviously Power BI, we'll bring that up, um, because it's kind of part of

47:08the ecosystem.

47:09Um what else have you got in Azure that really works well with Windows Active Directory, uh all that jazz.

47:15You've got Microsoft Suite, obviously.

47:18I these are not in Azure, but they are intrinsic in your contract with Microsoft, I assume, to have Office 365, Azure, and everything else.

47:28I think it's a level basically.

47:33I'm sure it's levels, yeah.

47:34Yeah, yeah.

47:35Whereas Amazon

47:37Apart from the compute and some of that database technology, which is pretty good, Redshift never really took off like it was supposed to, right?

47:44that Redshift had its sort of snowflake moment and then fell flat in my opinion.

47:48It never really sort of lit the fire of the industry because yet the Amazon pricing model just meant that

47:54the more you used it, the richer they got and you know, it didn't the economists didn't quite level off and actually we're seeing a trend to move away back to on premise now.

48:03Uh with a lot of companies becoming quite cold bare metal.

48:07Yeah, exactly.

48:07I love that phrase.

48:08It's bare metal.

48:12I just think like the the the the the the hardware engineers needed something cool still like yeah bare metal man anyway um

48:21The cloud path was super interesting.

48:23GCP doesn't have something that hooks you in.

48:25That's ultimately the problem there, right?

48:27And from a tableau perspective, all you're really doing

48:30connectors potentially some of the data sources in there.

48:34You've got uh is it Firebase, which is the really commonly used GCP um data source, I think.

48:42GCP, hold on there.

48:44Google this.

48:45Um of course with Tableau, I forgot I do forget it's not really GCP, you've got your connectors to um

48:56BigQuery, Google Sheets, Google Analytics as well, right?

49:01Those are like your big things, but Google Analytics and Google Sheets don't need Google Cloud Platform.

49:06I know they probably run off the same architecture, but I don't think you really need them.

49:09Um trying to look at see what else they've got.

49:12Not really much else.

49:13But Fire Ace is real time.

49:14They've got MongoDBs.

49:18Compute Engine VM, then download okay, analyst platform for your entire team.

49:22So you can put Tableau server on their compute, on their VMs.

49:27So that's probably the only other use if the Tableau server is not gonna die a death slowly.

49:33But uh yeah, like that's about it, right?

49:36Um run run run your

49:40server and containers all over in in in in Google.

49:43That's pretty much the end of that.

49:46Good.

49:47Honorable mentions before we call it a day.

49:50Uh we touched on Power BI.

49:52Don't really use it with Tableau, but I kind of think like if you're a day travelist today and you're only learning Tableau, you're kind of not really selling yourself to the market.

50:00You've got to be learning Power BI.

50:02And I say that in the nicest possible way.

50:03Someone who hates Power BI will not be made to learn Power BI properly, unless I really have to.

50:10I've done a good enough job talking about Tableau to avoid that.

50:13Um but yeah, like if you're P come at us.

50:16Yeah.

50:17Yeah.

50:17And there's actually, you know, Baz power there's

50:19There's quite a few Power BI channels.

50:21None of them went with my naming moniker, so maybe that's a hint I should have taken.

50:28So that's good.

50:30Sigma comes out from time to time, competitor of uh Tableau.

50:33Don't have to use it with Tableau, but if you were going to learn something super, super bleeding-edged and modern, it'd probably be Sigma.

50:40Where I think Sigma falls down, have you used it?

50:43I've just seen some guerrilla marketing, that's all.

50:45Yeah, so it okay on the marketing

50:50I'm just not a fan of Sigma Marketing.

50:52It just misses the whole point of the tool, in my opinion.

50:56Waste too much energy focusing on Tableau when they should just focus on the thing they're really good at, and that is enabling people to build analytical workflows.

51:03End of.

51:03Not workflows in the sense of all tricks, but workflows in the sense of dashboards that should be apps.

51:09Sigma goes that one level further and lets you build that app.

51:12right within the dashboard lets you go a little bit further than your typical thing it's got right back natively built in but here here's the kicker like want to connect to fat files not in sigma so it's also very

51:24Very modern.

51:25Like it's so modern, it might break some of the large, like you know, companies that we come across.

51:32Where unfortunately there is that one Excel file somewhere in the organization.

51:36That everything is reliant on.

51:38Yeah, yeah, exactly.

51:40Everything is in the cloud, this is in that.

51:42And then you go to you go and talk to the country manager and you go, oh, where's this lookup table

51:46Oh, it's in my OneDrive and you're like, you have that moment.

51:49Um, you know, it was like Chris Lawrence in Bad Boys, you know, he has that meme where like shit gets real and his eyes get bright open.

51:56You're like, where like why are you keeping that in OneDrive

51:59Yeah, happens in every situation.

52:01Sigma is not the tool for that.

52:03But I'm enjoying their sort of content.

52:05I'm enjoying what they're sort of saying.

52:06I love Luke stankies and um Luke stocks.

52:10You know, history of uh analytics uh that they keep doing on the Sigma channel, they keep doing these sort of short videos given given the heritage.

52:18Obviously, it's quite biased towards Sigma positive.

52:21But yeah.

52:23Interesting to see where they go.

52:25Stop focusing on Tableau.

52:26Just just just just do the same thing.

52:29Yeah, exactly.

52:30And it's doing its job well.

52:31It's just too focused on Tableau.

52:34But the other problem with that is Tableau is an actual platform, uh like in the nicest possible way.

52:40If I say

52:42Tableau, Sigma, Power BI.

52:45You just immediately discount Sigma to any customer.

52:48They've never heard of Sigma

52:49And so they've just got this really long path to building.

52:53Name recognition matters, right?

52:54Like name name recognition really matters.

52:57And it it's that age old saying no ever no one ever got fired for, you know, uh buying IBM, right?

53:03Like

53:04In analytics, no one ever got fired for buying Tableau and Power BI.

53:09That is where those incumbent tools are.

53:11And so

53:14Yeah.

53:15Anything we missed?

53:17Any other honorable mentions?

53:19Don't think so.

53:20I think I mean uh yes, th there are many that we have missed.

53:24Um but I I don't think there's anything major that we want to talk about that we said we would.

53:29That we want to talk about is a very fair way to put it.

53:32Like this I'm sure there's other tools.

53:33And if we've missed you a tool and you love your tool, we're really sorry

53:36MongoDB, like you can you can keep reeling off all of these things.

53:40You can do the game of tech tool or Pokemon, which I think is still that that that website is still live.

53:48But no, it's I think it's it's useful to just go through and talk about these things.

53:53And I think is this it will this episode be as dry as that hyper episode we did?

53:59Maybe for some people

54:01Um but I think I think it's you've not seen my editing skills.

54:06Wait till I get these into TikToks and shorts.

54:11Exactly.

54:12Yeah.

54:15That'd be filthy.

54:17No, I think it's um I th I think it we we've we've covered a lot there.

54:20Um but no genuinely if if you have, dear listener, any thoughts, anything that you think we should we should dig a bit deeper into or

54:28Uh you want to come on and discuss something with us?

54:30I think you know one of the reasons I think we wanted to do this is almost to open the arms of like, tell us we're wrong.

54:37Yeah.

54:37Exactly.

54:38Come talk to us about it so we can grill you on um

54:42And learn learn at the same time.

54:44Like I think, yeah, Enzo Databricks, maybe even some product manager, somewhere at Ultrics, who's like clinging on being like, please don't go.

54:53Um yeah.

54:55All good.

54:57Right.

54:59Next episode we should talk about 24.

55:012.

55:02Um we should talk about um

55:05Some other stuff.

55:06It's Tableau dying.

55:08Let's go back to the core, Andy.

55:09Let's go back to the core.

55:11BAM!

55:1224.

55:132 is Tableau dying.

55:15Does Salesforce care?

55:17Let's cover it all.

55:18Let's just deal with it.

55:22Hot takes to come.

55:23Hot takes to come.

55:25Those are those are great topics.

55:26We will like genuinely cover them, I jest, but um

55:30You know, I was in touch with someone on Reddit who posted something.

55:33Oh, is it even worth learning Tableau anymore?

55:35Blah blah blah.

55:37And I was like, that's the wrong question

55:40The question should be asked.

55:41Which parts of Tableau are worth digging deeper into?

55:45Exactly.

55:46Viz extensions.

55:47Exactly.

55:50Yeah.

55:51If it's the same thing that, you know, everyone's blogging about still from the last five years, no.

55:57Don't like it's not about building beautiful stuff in Tableau Public to become like, you know, top data analysts like

56:05The way companies are going, people are moving further into the stack.

56:08So we'll talk about that next week.

56:11And where I think like the sweet spot in terms of skills is.

56:14And also I'll say something controversial.

56:16It doesn't all have to be about Tableau.

56:18Like people shouldn't be learning anything wedded to a tool.

56:23They should be learning agnostic skills.

56:25It's easy to pick up the tools once you're in front of them.

56:29It's hard to know the core skills whilst you're in front of the tool.

56:33So learn those first.

56:35Anyway, more next week.

56:37To be continued.

56:40Take care, mate.

56:41Super.

56:41Take care, everyone.