Mass-format reports in Power BI - Ep. 369

In this episode, Mike and the team discuss Mass-format reports in Power BI—what it is, why it matters, and how you can apply it in your Power BI work.

News & Announcements

• Fix visuals, replace fields, and mass-format reports in Power BI — DATA GOBLINS — In this article, I explain several use-cases for semantic-link-labs with reports, which involves programmatic modifications to report metadata to streamline certain report management tasks. This article explains how…
• PowerBI.tips Podcast — Subscribe and listen to the Explicit Measures podcast episodes and related content.
• Power BI Theme Generator — Power BI.tips - The worlds best theme generator for Power BI reports. Increase your speed to develop stunning reports using this free theme generator. Themes are essential for any report developer’s tool belt. Visit…

Main Discussion: Mass-format reports in Power BI

The conversation dives into Mass-format reports in Power BI, including practical considerations, common gotchas, and real-world tips you can use right away.

Looking Forward

Try one small experiment this week related to Mass-format reports in Power BI—then iterate based on what you learn. Small, repeatable improvements compound fast.

Episode Transcript

0:36 good morning and welcome back to the explicit M podcast with Tommy Seth and Mike good morning everyone how’s it going good morning and a happy Tuesday to you to you gentlemen good morning happy Tuesday indeed Seth hey Mike you look a snazzy well we were traveling for work this week so I got to plan things a bit remotely I grabbed some lights off Amazon and boom look at that hotel room is now your studio so they should hire you Mike as a like how

1:07 should hire you Mike as a like how how to how to flip a a hotel room into a podcast studio in under $100 or something like that or under whatever that is budget budget video editing here in this one trying to go go go easy but hopefully we’ll hopefully you get at least decent audio and it still works okay for the the podcast so we’re trying to keep things going while traveling and doing things for work so anyways quickly jumping into our topic for today so our main topic for today is an article from

1:37 main topic for today is an article from our like we do this like every other month it seems like now another article from Kurt ber just working through Mass editing visuals inside power Bay reports how does that work what is the options here what can we do to edit things in Mass for our reports with that being said Tommy any news announcements that are coming out pretty big one so besides all the conferences and all the fun things happening Microsoft ignite coming out

2:07 happening Microsoft ignite coming out fabcon Tabler editor came out with their monthly update and they actually came out with a really neat feature there’s some other great enhancements like setting value filter Behavior but this new feature called code actions and really what this is is once you’re in the expression editor you basically rightclick you can basically choose either improve choose change the readability or reite

2:38 choose change the readability or reite it and these code actions are really like almost like macros but just actually in parts of the dax’s code or code or expression they’re incredibly easy you can also use your keyboard shortcut and you can actually create your own custom ones there’s some 29 customly built in like simplifying switch avoid the calculate shortcut syntax but you can actually write your own or or you can turn them on and

3:11 own or or you can turn them on and off so that’s a really neat feature so is this more about like formatting the code when you’re in the like so for example the examples they giving here are like you’re in the Dax expression editor yeah manipulating things and it’s giving you s like ease of use or common editing patterns that you do over and over again right right for example if you right click in the refractor it’s like hey remove all unused variables or you remove all unused variables or remove a table name know remove a table name split a multicolumn filter into multi

3:42 split a multicolumn filter into multi multiple filters so those are pretty down down KN interesting definely quality of life improvements right yes and and while I I I can certainly appreciate it I think we are spoiled if we’re going to look at how AI her you we’re going to look at how AI her some of the code editing assistants know some of the code editing assistants have improved performance of things like this is a welcome thing to see just in terms of like an assist like helping you

4:14 terms of like an assist like helping you build better code it’s not AI driven doesn’t seem to be but in terms of like de like simplifying the code editing experience you can wrap yourself around many axles and taxs right so it’s like hey clean up my code like take out all the stuff I don’t don’t need in my final set that that seems quite useful well for those of us who running like oodles and oodles of decks anything you can do to make it easier write decks automate things is like that’s a win for me yeah and I think they have a new

4:44 me yeah and I think they have a new tabl there has a new new like way to authenticate so I don’t know about you guys but I do a lot with the premium service like connecting to a a semantic model but a lot of times it’s like I have to log in again you have to go through the multiactor o on my computer but now you can actually change some of those settings that like basically I I believe like opening a browser so if you’re going connecting to different workspaces it still will remember you so little things like that because I really hate

5:14 things like that because I really hate writing Dax measures hate writing Dax measures in powerbi now now that now that you bring this up I wonder if this is something around the area of I think about what I do today in desktop and now that we have dax’s query View there’s a little bit of competition in the marketplace for like can we automate things in the Dax editing experience now and it’s nice to seeing other tools stepping up here and adding additional features here that are not included in desktop so I

5:44 are not included in desktop so I think maybe there’s something there because in in the desktop you actually have a a rightclick action where you can script out multiple Dax things inside the editor window inax quer R viw which I think that’s good it means there’s some like pushing of people to to make better UI experiences inside tools awesome speaking of added tools there’s a another one that Microsoft launched on fabric on the fabric update blog there is AI skill so co-pilot and

6:14 blog there is AI skill so co-pilot and AI skill in the data warehouse announced o Tommy should like this one that feels that feels very similar to like the data bricks experience where there’s a the the panel on the right hand side that and and incode editor right so if if your code’s wrong sometimes it’s just a you code’s wrong sometimes it’s just a right click enter rerun with the know right click enter rerun with the new code or explain it on the right hand side regardless of the tool sets right

6:45 side regardless of the tool sets right like this it is probably the the most used feature I I use that that accelerates code development for me but this is this is so just to be clear this is a The Experience here is code editing sequel with co-pilot inside the data warehouse inside Fab yeah yeah oh okay see this the co-pilot stuck does really well when there’s code involved and that I love where they put it there I I don’t really dig the fact that I’m trying to

7:15 really dig the fact that I’m trying to push copile it out to like build report pages and like other things but like SQL python notebooks oh my gosh it’s so helpful to have it there I sometimes being the pessimist especially on this show is is great I don’t know if I’m right but I like being right can you Ma imagine a world for a moment right like chat GPT I think opened the world’s eyes to just how powerful Ai and llms can be right yes but in terms of how how we’re getting

7:48 but in terms of how how we’re getting the most value out of it right now and this is also speaking from like a guy who reads A Lot knows a little bit about how we’re building these AI mechanisms but is not an expert but if I look at it incrementally the win to me should have been like holy like we are transforming developer experiences right we are accelerating this stuff so fast across all these ecosystems because you’re leveraging the structured nature

8:18 you’re leveraging the structured nature of how we develop code right yes you you have the thing there already but the minute you go from like these engines that have access to all data right which is how chat can be good although I did see an article segue parking lot I did see an article about like whether or not whether not that’s going to be remain to be valuable and I’ll tell you why in a second but back back to my point is think about that launch AI into these like Progressive steps into the org as opposed to this broad stroke thing where

8:49 opposed to this broad stroke thing where like AI is going to solve all the problems in all the formats and I think that’s what companies are struggling with now because as you’ve launched this out there and we’ve tested a lot of that regardless of what company is is deploying it yes there are some value ads but it’s not it’s not as significant I think as what we see in

9:09 significant I think as what we see in development Realms interesting another area I was just recently talking with someone that I thought was very relevant here in this generative space one area that is very in need of generative information is people writing like advertising things advertising agencies because they’re always looking for like okay here’s the context of what I’m going to be talking about I need three bullet points or I need three catchy phrases they’re going to like write this and I find that well and in talking with

9:41 and I find that well and in talking with them they’re training their team they’re sending their team to corporate to go learn how to best use AI skills from AI engines in their business and I think I was like a little bit taken back I was like okay I can see the value but I was like the bigger point for me was organizations are now saying hey AI is going to change how you how you do work your companies need to identify where is a good place for AI to be a place and spend money on training your team how to do better AI yeah to make themselves

10:11 do better AI yeah to make themselves more productive this is going to be a thing you’re going to go to AI training now right it’s but it’s it’s the tools too right like AI is a tool and I think I think the way it shapes things more so for like like you’re saying like marketing or it can generate IM generate videos like it’s it’s a create like humans are needed for the creative part of it yes right you’ll always be what is interesting for some of the LMS right like I was reading an article

10:42 right like I was reading an article in LinkedIn I wish I remember who wrote it but the hypothesis is at what point in time do tools like chat GPT or llms and I don’t know if this is this is true or not but like it’s an interesting thought exercise if they’re sourcing all of their their answers and information from places where humans are solving these problems right like stack Overflow and like all all the other communities if people stop using those communities

11:12 if people stop using those communities and only start using chat GPT then you don’t get any of the new stuff in and is that going to impact or affect the outputs of those llms I thought it was just a very hilarious like cyclical thing if it if if it true came came about to be it’s funny you say that I just started reading a book this weekend is by a guy who’s like one of the maybe the smartest person on like prompt engineering and I’ll put the book name in there because I don’t remember the book name but he’s talking about

11:43 the book name but he’s talking about really the history of LMS in that that big jump we went from I think remember how Microsoft they they tried like two AI tools and they both were terrible back in the day I’m not going to name the names think but one was a Twitter bot that basically everyone it did exactly what you did the internet basically would tweet it and say back whatever to her basically and it was down in 16 hours and then they did something in bing before

12:14 they did something in bing before co-pilot and one of the one of someone from like Forbes or the New York Times realized that thing was trying to hit on them basically so to speak so because a lot of the the the content that llms have been getting is public domain right so like how do you actually label yeah exactly but it’s like you you you look at and why I think that idea has legs right is because if you look at the large language models that have

12:44 the large language models that have access to everything they’re they’re very informative you can get some really good ideas and answers about a lot of things but when you open those in smaller lockdown ecosystems like your company because you don’t want your company data to be out there what are we find what are we finding like they’re they’re not the it’s not a one text box ask anything in your own internal Ai and you automatically get the answer right so this is where we’ve had many discussions around like there’s

13:14 had many discussions around like there’s got to be foundations of data there’s got to be some structure to the data like why are they so good in the code editors because there’s there’s bounds there’s Cod like you can only do certain things right and if you’re learning the bad patterns and recommend in the good ones or recommending like how the pattern of the code is being developed and the like the joins of the different columns and you’re just Auto recommending the whole join and all this stuff like that makes a ton of sense versus not having any bounds

13:45 versus not having any bounds but your bounds are this very small box right and and that’s where even on the larger scale if we’re not continually feeding the monster right like the the engine with new things at what point it does it need to start creating the answers itself right yeah so I don’t know this is I was talking with someone about that exact issue and they said basically came down to there is the ability for people to need to build new

14:16 ability for people to need to build new packages new code new things right if those things are not being referenced as frequently you’re you’re gaining this knowledge around amassing old technology because it’s been out it’s very common it’s very well documented that’s where things are getting trained on so new things that are more efficient or different or implemented in a slight different way they’re going have to think about how to train again this is all going to be an evolution of things I think they’re going to figure out what to do and they’re going to figure out hey this content’s newer let’s go

14:46 hey this content’s newer let’s go recommend things from there more frequently than things that are older there’s got to be something there in in the language models but they’re they’re interesting and I’m not sure we all know how to deal with them yet it’s touch and go I I guess a bit what’s up how do we Implement them the best right I think that’s where I think that’s where we’re at I I’ve seen a lot of success in the code area and creative spaces but it’ll be yeah it’ll be interesting to see how it evolves for sure but regardless I’m very pleased that there is a co-pilot showing up

15:17 that there is a co-pilot showing up where I’m going to be writing sequel at the end of the day like I’m I’m okay at writing sequel I can do most things there always weird things I miss it’s synta every time I forget the stupid group eye when trying to aggregate things every time like the the AI should just figure out what needs to happen and be like hey I figured out there’s you be like hey I figured out there’s and there’s there’s techniques in know and there’s there’s techniques in SQL I don’t use like who uses rollup like I don’t I don’t use it as often as I probably should but there’s other functions that might be more effective at doing things and having an AI

15:48 at doing things and having an AI guidance helping you write those things out it just makes things faster anyways I I’m a fan of it I’m very I did not know that came out and so I’m very pleased that that’s coming through to the data house that’ll be a lot of fun any other news items before we get into our main topic I think we’re just about at time all right with that let’s get into our main topic so today’s main topic will be the article from Kurt bu talking about fixed visuals replacing fields and mass format reports inside

16:19 fields and mass format reports inside powerbi now I want to I want to make a note here I think this is something that I identified when I was I in conference in Europe I made a huge statement and I’m probably I think I’m going to stand by my statement I was I was watching Michael kovalski write notebooks and building a whole bunch of things in reports with semantic link labs and I was thoroughly impressed with like oh my

16:49 was thoroughly impressed with like oh my gosh this tool does it does editing of a model it can do report editing it can do a verac EXT extraction it can then write Dack queries against the model it can use all the powerbi apis based on the context of the user and I thought holy smokes this one tool semantic link Labs is has the potential to make every other external to Obsolete and I tweeted it I I threw it out there I threw out a tweet and lo and behold

17:19 I threw out a tweet and lo and behold tabular editor team Daniel came in and said hey are you sure about that is this are you sure like there’s still probably some use for like Pro tooling but the more I use semantic link Labs I just did a workshop with for training. tips around semantic link Labs it’s phenomenal like it does so many good things and so I’m very pleased to see where this article is going to go and unpack this with you guys because I really do think the next tool that people should really substantially start

17:50 people should really substantially start learning is semantic link Labs I think it’s going to be a game changer I think it’s going to help you out with a ton of automation in talking with people they’re finding a immense amount of values for this thing so I’m very excited to see where we go with this article today just just putting that out there I I I’m going to stand by my word I think I think this is a GameChanger this is a major game changer for us building things in powerbi and and fabric for that matter it does it does more than just powerbi think which is

18:16 more than just powerbi think which is crazy Co we’ll see if we’ll see if if Tommy can keep you on topic with this one because there’s a there’s a lot to talk about with semantic link Labs oh it does so much stuff specifically today we’re it’s a timesaver right think the the article is is more around time saving in viewing copying and modifying your power VII reports with a lot of the different objects on all the pages right and I think Kurt brings us brings us up as far as an intro is concerned

18:48 up as far as an intro is concerned with all of the tweaks like right you you build a report you go through and then it’s six pages long and then somebody sees it and they’re like yeah it’s good it’s good but can we make it like 12o font and can you make this color this one and it’s the sh it’s the to-do list after you present it right that yes or or as another one that doesn’t happen as often I think is migration migrating from you I think is migration migrating from one version of something to the know one version of something to the other but there’s there’s a lot of

19:20 other but there’s there’s a lot of tedious steps to do in there especially if you’ve if you’ve so incorrectly not used a theme in in all of your stuff powered by of course the powerbi tip steam generator but let’s assume for a moment that you haven’t this is this is I think one of a really amazing way to interact with and and update that stuff albe it caveat unsupported a lot of this is right but we’re okay with that we we use pre

19:50 we’re okay with that we we use pre preview features all the time yeah I’m G to I’m going to push back on you a little bit on that unsupported feature where we are today there’s a whole lot more supported than what was in the past and I think maybe what I would put a caveat around this one is the underlying technology that’s doing this for us is the fact that we now have the pbir format powerbi report format which is totally changed how we store data about the report it it’s it’s that

20:20 data about the report it it’s it’s that is that needed to happen for something like this to exist which is I think my main that’s a really good point there yeah and I’m I’m not suggesting that doesn’t work but like Kurt Kurt even calls out right writing back to the report. Json of a powerbi file is file is technically still not sported well anything under the pbir format is in preview so it’s like you’re your mileage will vary and I’ve been having bugs with like I’ve been making PBS in and the old

20:52 like I’ve been making PBS in and the old original format because the pbrr has gone through an evolution and there’s some weird bug right now that I’m finding where if you go from an old PV format and try to open it again in a newer version of desktop desktop says hey you’re in the old version of a PBR we’d like to update you and when you update it it just breaks desktop it says I can’t save the file your work is not saved I can’t save anything it just like freaks out you literally have to close the file start back up and say don’t upgrade the file so there’s some they’re still trying to figure out some Kinks

21:23 still trying to figure out some Kinks and there’s some bugs I think still there I’ve got an issue to Microsoft to like let them know okay okay I’m finding some weird behavior I’m not sure if it’s just me making some bad stuff which could totally be a thing or something is actually wrong when they’re trying to update these files so we’ll see regardless everything I see though is I’m hoping in the next couple quarters or months they’re going to land on the PV format it’s going to come GA General availability and we’re just going to go to town on this thing because this is this is the right way to go yeah so Tommy what what

21:56 right way to go yeah so Tommy what what areas do you want to focus on here yeah a big article yeah and and I know that this is probably going to be a few part series because Kurt actually has three articles on semantic Labs report data set and then all over to me this is absolutely crazy and I want to think about the use case or the practicality as much as semantic it’s absolutely incredible he’s doing everything in a Jupiter notebook here to modify the report first he can

22:26 modify the report first he can actually view the report in a notebook and there’s actually isn’t there a Jupiter extension for powerbi where you can embed your own data but this is actually looking at yeah I actually I used it this weekend but we can actually launch a report in a Jupiter notebook yes which is pretty crazy to be so not having to go anywhere I don’t know guys with fabric I have eight tabs open of fabric now that’s just how life is I think I think

22:57 just how life is I think I think there’s a guy I talk to him like twice a week and all he does is talk about notebooks who who’s who’s that guy who Could That Be I don’t know guys you guys know who I’m talking I think I bring it up once a week you’re but you’re right like notebooks are are the way to interact with data they just and and and this is just like starts to pop your mind with

23:28 just like starts to pop your mind with just amazing capabilities of like how cool all of having all of these capabilities in these notebooks is it just blows my mind but floor open mike why should everybody use notebooks well I notebooks well to be to be to be frankly honest I mean to be to be to be frankly honest I mean to be to be to be frankly honest the notebooks are a crazy different mean the notebooks are a crazy different experience than what you’re used to developing with they’re awesome and I think Kurt’s example here in the first part of the article is actually really solid example like know

23:58 really solid example like know you build so one of the things I saw Michael kavosi demo was okay let’s imagine you have a semantic model and that semantic model has many reports that are attached to it one of the things you want to do is you want to update a measure name or delete a measure and you need to know what is the impact of that measure centic link Labs has the capability of saying here’s the model give me the list of reports and it can evaluate the impact of that measure change on everything Downstream now I

24:28 change on everything Downstream now I think kurtt misses a think it’s a very strong use case but Kurt doesn’t really describe that one specifically but in Kurt’s initial example he says here let me show you or with a couple lines of code pretty easy let me render a report immediately you say oh look a visual is broken because a Field’s wrong or something’s not correct on the visual and so immediately he’s able to go look at the report and say hey this is a problem there’s a field that needs to be fixed and it says there’s something wrong with one of the fields

24:59 wrong with one of the fields customer plan roll table the column called key account whatever the thing is right so there’s there’s a a message box that says there’s a problem and then he goes through in what is this like six lines of code not even he doesn’t even need to import semantic Labs but with only three lines of code he’s un able to spit out the PB format of the report it goes to get the report Json so describing the entire report back to like this this is crazy

25:29 report back to like this this is crazy and it’s all in a notebook and you can automate it this is again I’m just floored by how good this experience is so I really like Kurt’s initial example here I think it’s really solid what are your thoughts I’m just wondering I get the data part and maybe I’m G to push back a little at least with this article because I get that from data management data engineering I’ve been doing a ton actually just working with a semantic model or just table of data using and

26:01 model or just table of data using and it makes a ton of sense but to go through your report and to just to try to go through the the actual content of a report wouldn’t you want the user interface you want the UI than the canvas and give me a reason why I should never have the canvas again well I don’t think that I don’t think that’s the answer I think the answer here is like look I’m thinking of this example that Kurt’s doing is your going through reports like you have a lot of reports in your organization some stuff may

26:32 in your organization some stuff may be broken some stuff may be not or maybe you’re making a change to a semantic model that is a is a solid CH like you’re literally going to make the business requirement have adjusted you’re going to need to rename a measure that will that will 100% break things Downstream no matter what so this is a nice easy automated way to say look I have made one change in my semantic model I’ve got five reports to look at I can Loop through every single report and just say let me just say if it looks so good let me just look at it visually and you can go through this to me this

27:02 you can go through this to me this notebook in this example here this is part of your QA process your QA team or whatever look at the reports are going to use this to figure out what’s going on the neat part about this is that the notebook actually goes through and completes the fixing experience as well but it’s just one it’s just hard to identify where are the problems so this is again I really like this and

27:24 so this is again I really like this and now that I have to me this brings it full circle right make a change to a model identify where the change breaks things go look at the reports verify that it’s working again see those changes and then an example here he talks about fixing those changes so he’s fixing them adjusting the data properties of that Visual and then you can push the changes back into the report which is crazy cool and I’m less worried about this this is almost like this is editing the report in the

27:54 this is editing the report in the service so the report could potentially be different than what you have on like so for example if you if you so this is this is eding report in the service so this is like the equivalent of like hey I’m taking reports down in stadium and SharePoint because that’s how we store versions of our ports this is not that this is totally different but I think now with get integration on the workspace if I’m making modifications on the reports and to me I think the Lynch pin for me is in order to edit reports in the service I’m now comfortable with

28:24 in the service I’m now comfortable with it because there’s get backing it if there’s something really going wrong I can restore the old version of the report I feel comfortable editing stuff inside par. com without worry of not being able to download the file again sorry I said a lot of things there I’m not sure this all Mak sense I think it did it did okay I think what’s I think what’s really interesting to me is the a as semantic Ling Labs starts to expand and show like how the community is how you can leverage it your your

28:59 is how you can leverage it your your one of the most impactful things I guess to me is you’re keeping everything or bringing it all into the same interface which already is a huge timesaver right yes to your point if I’m in development making changes pushing those changes out from the same place and then validating and then fixing right instead of how many browsers are are open right like I wonder which which brings me to a a random thought we we I think have grown to the mega screens of

29:30 to the mega screens of computers I wonder I wonder if that’s going to start shrinking because now we can do more in smaller places but like in today’s world with powerbi like I have tabs open all over the place oh if you have an end to end flow going they’re like you’re you’re utilizing all of your screen real estate 100% you are what so what’s cool about this to me is just that that has struck me as I’m looking at Kurt’s examples and walking through that interface is I can just pull it up right I have this one

30:01 just pull it up right I have this one place and now I have hooks into other parts of my system that allow me to not just visualize but also remediate and then like hey we in the future I how far off are we from understanding which visuals are broken and automatically fixing them right or just doing prompt fix through all the the changes that were deployed yes this is where I think things I think I’m a half glass full person and you guys probably laugh at that because yeah Mike’s probably optimistic in nature but

30:33 Mike’s probably optimistic in nature but like I while it’s not all built yet I just see the potential of this feature or being able to use notebooks in this way this fashion just opens up a lot of doors and and a lot of the so we we built the business Ops tool for powerbi so business Ops helps you manage external tools to your parbi model I think with these semantic link labs and particularly the notebooks being able to so easily manipulate and edit things against the service I think some of

31:03 against the service I think some of these tools that we we put into the business op tool are a little bit meaningless now you can read all the measures and format them all at once you can edit reports now like there’s there’s things there was tools that were being made that like one-offs to me this is like a single place where we can do all those features in one spot so I’m excited to see where the community continues to automate around reports and what else can be built there I I think there’s going to be with semantic link Labs I really feel like this is the starting point for like

31:33 like this is the starting point for like really true data Ops right checking your data confirming that it’s right making sure that it’s it’s correct in an automated way so do a deployment make your model push it out the test run a bunch of tests with semantic link and I think I think this is where things get really exciting for me because I think that’s going to really improve our quality and speed up what we do I think that’s a huge win I just I just want to point this out I want to give a a big Kudos right now

32:03 want to give a a big Kudos right now which we’re at 369 episode have I ever done this before maybe first time this is a first time for everything huge Kudos not only I think to the community for sure but also the decision makers at Microsoft to go down this path to open up the the PBX file and powerbi because had that not happened this would not be happening we we would not have this because it was such a black box

32:34 this because it was such a black box that you couldn’t do much with and couldn’t interact with and because those decisions were made previously we’re we’re at this amazing point where we’re talking about powerbi in a notebook being edited and being able to push change like just that is so cool to see and I’m I’m like I as you were talking and as we’re like know talking about this like I get those bubbly feelings like oh man love it when love it when those things that are

33:05 it when those things that are incremental and you don’t see the impacts this is one of those like impactful moments to me where it’s like dang like this is this could be a fundamental change with how people and organizations interact with their reports so a big part of this though is the pbir file and I think I don’t know if that’s that is the tech yeah that’s the tech that’s what I said earlier is like every all this doesn’t happen without the PBR and and I think this is a pretty big thing where everyone was so used to and

33:36 thing where everyone was so used to and happy with their PBX file and this one we weren’t happy with it it just was what we had to use it what we had right it was I get it it was It was built one way this is but this is what happens though you don’t know what you don’t know right there’s a lot of tech de they had to like figure out what to do to make reports work and so do we build the PB format day one probably not like we just build something that works and then it works and then you come back and say okay how can we build it better this is the optimization step of building reports like this is optimizing what

34:07 reports like this is optimizing what they’re doing making it easier there’s a in the PBX there’s a bunch of stuff in there that they’re not keeping in the PBR format it’s way trimmed down there’s a lot of features and settings and properties that don’t exist in the PBR because they don’t need it they’ve figured out how to work around it but that’s removing a bunch of tech debt from that file format that is now not required because desktop and the service just knows how to work with it so right this is just them learning and doing a better job with what they’ve learned about I just I’m super thrilled about

34:37 about I just I’m super thrilled about this one one thing I’ll also note here this is a maybe a general observation on things and and this is maybe a feedback for Microsoft anytime you can build a a tool or product that involves the community of passionate people to build Solutions or find problems and fix those problems with tooling that enables the community to build something right every feature that I think has gone wild inside desktop I’m thinking external tools I’m

35:07 desktop I’m thinking external tools I’m thinking the xmla endpoint opened up to external tools to be allow them to edit stuff like when things get open sourced or lets the community involve like here’s some automation we’re going to give you access to the standards that Microsoft produces that’s the time when things get like really exciting like I get really excited about it people build cool things some of them extremely useful some of them Me Maybe not so much but it’s interesting but I this to me this is like the more you can think about how do we put Community

35:37 think about how do we put Community First in whatever we’re building I think that actually accelerates the usage of the product and so maybe I’m missing something here maybe that’s not the goal of Microsoft ultimately and I’m just picking up on a couple key features that I like but I think the takeaway for me here is whatever you’re building whatever application tool software think think about how your community can enhance it or give them you you build this the standard of what it starts out as but then figure out a way to like let the community invest their time and

36:07 the community invest their time and effort and now the community is building stuff for free they don’t Microsoft doesn’t pay them to build more notebooks and examples of this they’re just doing it because it’s interesting anyways that’s just a general observation and it it drives more usage and it’s a benefit right exactly so so random question right ear early early too early to tell and I but how how far

36:31 too early to tell and I but how how far do you think this one potentially takes off in terms of two two parts I is this does does semantic link labs and this interaction like as it continues to evolve and grow it become like one of the defao ways that like the hardcore powerbi developer works right and and or does it does it does it have the capability to stretch outside the bounds to the Del lay powerbi user or

37:02 to the Del lay powerbi user or Not O this is a good question I I I didn’t think this was where you were going with it one with this question I think in general what I’ve observed around notebooks there’s a large hesitation from the powerbi let’s just talk pure powerbi Community right pure power there’s a large hesitation from the powerbi community it’s like ah it’s notebooks it’s fabric it’s writing a lot of code a little B it’s very technical y so I think I think there’s an initial blush of blush of hesitation when I think about what

37:33 hesitation when I think about what has really worked well for people is when you have that DBA level person the one who understands the data the one who’s writing SQL statements the one who’s writing store procedures those type of individuals don’t have that much of a gap to jump between what their knowledge skill is and in using notebooks there’s definitely a learning curve for sure but I think the more I work with companies and the more I teach them about notebooks it doesn’t take long before after a couple patterns

38:03 long before after a couple patterns people just start getting it things start clicking on and they’re like oh this makes a lot more sense they’re saving their code they’re finding again it’s some of these things are like I don’t even know what to Google or search for to get the answers that I need right if you get co-pilot down to these lower SKS inside Fabric and users can start asking coalite hey how do I write a class statement in Python it just shows up it just shows you here’s an example of one oh that’s interesting and so when when that

38:34 interesting and so when when that starts when when you start bringing that level of like assistance down to those new powerbi users I think they love it on the flip side of this there’s another user group that I think is the dbas they are already like very technical already I don’t think they’re having any hesitation honestly I think they’re just starting and going moving forward with this one so my opinion is I think it’s going to take some time I think having good community support building examples and documents and articles but of the things I’ve seen recently everyone is

39:05 things I’ve seen recently everyone is writing articles about notebooks these days it feels like there’s just non-stop content coming about notebooks and how good they are so you think this takes over the world Tommy what about you I don’t because as much as notebooks are awesome and like I said it’s been a amazing learning curve for me or kind a amazing learning curve for me or like a journey on there for the last of like a journey on there for the last two and a half years where it’s something I it’s a go-to granted there sometimes I’m like it’s easier in power query but I digress I I think for the

39:40 query but I digress I I think for the the the mass users they’re not going to start trying to Learn Python and maybe their organization has co-pilot with the jupyter notebooks already enabled but I don’t know about you the co-pilot notebook feature it’s like a help feature but it’s not necessarily like here we’re going to llm on that code that you have actually given if I think if honestly if if Microsoft can integrate away with like a VSS code or

40:11 integrate away with like a VSS code or something very simple where the co-pilot and GitHub which is probably the best code assistant out there where they’re like I want to connect to my data set done can I write the code for you is going to be a good way because there is a pretty steep intro compared to power query where power query is easy to learn compared to user interface is all all right there right all right there and compared to

40:43 right all right there and compared to the different data types in Python you have to now know like P I wish power query had packages but every you power query had packages but every all the different packages where do know all the different packages where do I get help for those so I think that unless I know I’m going to be working in Jupiter notebooks on a daily or weekly basis a person’s not going to just begin to pick that up got but this is my this is my point though my point is you’re not going to naturally pick it up but I’m telling you like from one who knows how to like look

41:15 like from one who knows how to like look at stuff and who’s learned some things like this is worth your time to learn that’s my point my point is like you may not be like interested right away initially but the sooner you just jump in and say this is worth learning and you start spending effort it will help you it will make some things more easy fluid simple to work with I I just I’m gonna push you to like this I’m G to push people in this one I’m like the powerbi user are with you Tommy they come from Power query they like all the buttons it’s easy to click the transformation just happened great works

41:45 transformation just happened great works well but right now man if I’m looking at things the data bricks sorry the the notebooks team is eating the power Queen the power query team’s lunch at this point like it the stuff that is coming out from from the the spark and the notebook side is just way faster cheaper to run easier to work with and features are coming out really quick in the space now maybe there maybe Microsoft’s playing a lot of catchup here but like dang this is it’s really cool this is yeah you got to learn it regard

42:15 this is yeah you got to learn it regard regardless I fall I fall a little bit in between and and I I think what’s exciting to me about Technologies like this or features that allow us to SC SC and just continue to are the these are accelerators to workflow right and the more the more tools allow us to accelerate the more adoption you’re going to get right because at some point in time like as opposed to where I think

42:46 in time like as opposed to where I think we’ve been right like specific tools allow you to accelerate on One path right but when you can add multiple things together right so now if I engage in understanding how this works I have a 5x multiplier or a 10x multiplier like now you gain the interest of a wider audience so I think Developers for sure like longterm like this this is the route right and I think more and more

43:17 route right and I think more and more will adopt and and learn a lot of the the Frameworks and new things you have to to to to get it all I would say from the outer fringes of like engaging with the powerbi users I I’m with Tommy not everybody’s going to adopt every like you’re you’re going to have a large spoth of people that need the UN user interface they they’re just like Excel right I want to control cell byell I want to see that my data in these things

43:48 want to see that my data in these things that’s how I understand it my LEAP was into Power query where I can finally understand my steps I can see them all it makes sense because I don’t have a technical background however I will say that tools like this allow the business user who wants to engage in those time-saving activities the stretch is much less I think right like because you you open

44:19 think right like because you you open this door to creating business rock stars where they’re like and and I think this resonates with you Mike back in the day because you’re like I’m an engineer I’m lazy I lazy meaning I want to automate all of my right yeah I can and where notebooks I think are this amazing tool is like you have markdown you can basically create tutorials of like how you go about doing something doing your work yes and and and from the technology

44:51 work yes and and and from the technology side from the developer side I think it behooves you to almost say like oh okay well how many of these tutorials could we build that do these things that do things for business but also at the same time are training them how to use it right and I think that’s the max value you get out of this experience because it’s not just okay go read all my SQL code right it’s markdown here’s what we’re doing here’s the comment that’s the cell this is what you

45:21 comment that’s the cell this is what you load if you want to like read the cell go ahead but I’m not breaking your flow from top to bottom or anything like that so I think I think there are more opportunities to reach out into that like to broaden the audience a bit because of that notebook experience but yeah it’s it’s going to take them

45:41 yeah it’s it’s going to take them the like the time and effort more so than to to leap into a deeper pool than they have in the past let me give you one other yeah love that point Seth and I want to come back to to the article from Kurt I think there’s a couple gems here at the end that we didn’t talk about at the very end of the report just before his conclusion he starts talking about some things he goes through here and he goes look I’m going to fix the visual okay cool interesting there’s another section here he goes there’s another whole section around and this is where I’m like this is the game changer that we’re

46:12 like this is the game changer that we’re waiting for and this this is the part take note here creating reports from an existing report. Json so he’s taking an old report or report that he’s already built and he’s using a new command create report from report Json in there these are the only PR you shove in you shove in the report name the data set name the report Json the definition of the report in Json format which is what it can read out anyways the theme Json and the workspace where you want to

46:43 Json and the workspace where you want to put it one of the things that we have had issues with with the cicb pipelines the plet pipelines there’s been no high amounts of automation between hey I’ve got a report in test or Dev I’ve got to move it to test and I’ve got to rebind that report to a new data set and a new thing and so you think about all the deployment patterns that companies come up with like there could be a number of things you could have five six seven environments you could take one report from a central workspace and deploy it to a lot of like child workspaces

47:13 to a lot of like child workspaces because there’s things for HR and you want to maybe you want to duplicate a report between two different workspaces just because it’s the same report two different audiences like whatever like this part of the notebook he just he literally it’s like two paragress at the end I’m like this is the more important part of the article here in my opinion this is the landing of the scenario like this provides CI continuous integration continuous deployment with reports where you can start automating the movement of them and rebinding them back together again

47:45 and rebinding them back together again this is this is but this the only reason this is possible is because we have API this is what it’s doing it’s nothing do anything magical it’s using the powerbi fabric apis inside some link to go do Automation and to me like okay yeah cool this is like interesting things but like this is underrated like this is amazing and you’re going to be you’re going to see I mark my words you’re going to see a lot more examples I can’t believe John kki has not jumped on this bandwagon so calling John kky if you’re out there

48:17 calling John kky if you’re out there U please write me some more notebooks around Automation and moving things between workspaces using only semantic link labs this is amazing and I and I think what’s Happening Here is if I had to put my thinking tap on here when Michael kavosi was talking about semantic link Labs he talked about it in a way of like the labs is experimental stuff that Microsoft is trying to figure out what works of the things that work the best those will get pushed back into semantic link so Labs is this experimental place where they’re

48:47 is this experimental place where they’re figuring out things that are like it’s the early version It’s the preview it’s like let’s see what the community builds and then from that they’re saying look these are the features that are getting used the most we’re going to push them back into regular semantic link and now to have take you you’ll have this ability of like doing all this full automation like it’s just it’s incredible so it’s it just it just shocked me like when I saw all this and I started having these Revelations I guess of like this is a

49:19 Revelations I guess of like this is a real tool like this is something to take note of anyways I’m just really excited about it shocker Michael legs code that being said and notebooks are super awesome anyways I just wanted to point out the very end the very end in right before the conclusion was like a gem that I think that people you’re going to need to learn it’s going to it’s going to change how you do your workflows in the the future any final thoughts Tommy you have any final thoughts wrapping this topic up for today I’m really excited to get more on the semantic links especially

49:50 more on the semantic links especially what you can do both from looking in modifying your tables and even what you can do with your already your semantic models and I think that’s going to be a great use case I I’ve been saying this for a while when we think about the data data science aspect of this and then extending the semantic model that’s already has everything predefined to keep that source of Tru so I’m really excited to see where this goes do you really see a lot of customers ask I’m just be honest people talk about data science and how this is going to

50:20 data science and how this is going to help get data out of models and things like that do you see a lot of customers actually doing data science and using the needs to go after these semantic models are they asking for this yet or is this more of like a a nice to have feature and well eventually we’ll get there see this is this is what I’m going to make the arum they don’t they’re not asking for semantic models because they don’t know it but they’re asking for sources of truth because they’re like hey I just want to you guys are working on your data engineering data we’re working on this RW data but

50:51 we’re working on this RW data but there’s no there’s no inter interconnection they’re not thinking I want a semantic model that already has everything pre to find it’d be great but my question is more about the data scientist though my questions around I’m Challen I’m challenging you a bit on yeah no I see what you’re saying I I don’t here the story of the data scientists are showing like this is all bi business intelligence getting reports out the door to users in in the business I don’t see a lot of data scientist yet showing up to like knocking at my door and being like we have to use fabric for data science at this point it it it feels like it’s a far off story used

51:22 feels like it’s a far off story used probably in certain scenarios but not where I’m seeing it heavily used it at this point that’s and like oh my friend that’s exactly what I’m saying I don’t think right now data scientists are like thank goodness there’s semantic link Labs but I think there’s going to be a I think we’re actually gonna have a story here where the semantic model is going to be a end point or connection point for a lot of what data scientists do got it okay Seth what any final thoughts I’m

51:53 it okay Seth what any final thoughts I’m I’m excited about the direction here and I’m I’m I would imagine and we talk more about sematic link labs and and the features that are coming out these these to me are like a growing number of use cases that as the community continues to show them right and show others what you can leverage with this tool these These are these are big timesaver kind These are these are big timesaver Articles and tools and of Articles and tools and those those get me those get me excited

52:23 those those get me those get me excited so looking forward to where where it heads thank you Kurt for the article so always a great one and good topic good talking points all right with that being said we just appreciate your time we know your time is valuable You’ spent a good hour with us just talking about code and semantic link Labs we hope this has inspired you to explore a little bit more around what this is for sure go read Kurt’s article it is in the description of this video below go read it learn what he’s doing the code is simple don’t worry it’s not too complicated you can at least follow

52:53 complicated you can at least follow along with your own reports and things in your environment as well so I think this is going to be a win for you this is this is going to change how we build things in the future this is this is going to be a fundamental tool that everyone will be using moving forward and the best part is it’s free it doesn’t cost you anything to use it there’s no licensing it’s just go install the library and it just works with inside fabric which is Again The Price is Right if you’re going to if you’re going to go get wide adoption

53:23 you’re going to go get wide adoption so that being said if you don’t mind if you like this so you thought this was valuable to you please do us a favor subscribe hit the Bell that helps us get the video out and then also let somebody else know that you enjoyed this video and hopefully they’ll jump in and listen as well Tommy where else can you find the podcast you can find the the Pod you can find the podast on app wherever get your podcast make sure to subscribe and leave a rating helps us out a ton do you have a question an idea or topic that you want us to talk about

53:54 or topic that you want us to talk about in a future episode head over to power tips podcast leave your name and a great question finally join us live every Tuesday and Thursday a. m. Central and join the conversation on all of power byad tips social media channels this is how it’s live when there’s there’s a couple like flubs in the middle of we trip we don’t we don’t always get it right all the time but sometimes we trip over our tongue excellent thank you all so much and we’ll see you next time

54:24 and we’ll see you next time [Music] [Music] out out [Music]

Thank You

Thanks for listening to the Explicit Measures Podcast. If you enjoyed this episode, be sure to subscribe and share it with a colleague.