Fabric, AI & The Future of Consulting – Ep. 309

A new episode of Explicit Measures is live—this week the team dives into Microsoft Fabric, AI, and how it’s reshaping the future of consulting and BI services.

News & Announcements

• PowerBI.tips Podcast — Subscribe and listen to the Explicit Measures podcast episodes and related content.
• Power BI Theme Generator — Create and download Power BI report themes using the PowerBI.tips theme generator.

Main Discussion

This episode digs into how AI (and Copilot-style tooling) changes the day-to-day reality of analytics work—and what that means for consulting and internal BI teams. The big idea: AI can accelerate output, but it doesn’t replace the need for clear definitions, good modeling, and governance.

Key points from the conversation:

• What consulting customers actually buy: speed, expertise, and risk reduction—not just “hours of building reports.”
• AI as an accelerator: where AI helps (drafting, exploration, boilerplate) vs. where humans still have to make the hard calls.
• Quality + trust still matter: the semantic model and business definitions are what keep AI-assisted work from producing confident nonsense.
• New skills for teams: prompt literacy, review workflows, and the ability to validate outputs become part of the job.
• The future role shift: less time on repetitive mechanics, more time on architecture, adoption, and decision support.

Looking Forward

If you’re introducing AI into your BI workflow, set review standards and model governance first—then scale the productivity wins safely.

Episode Transcript

0:31 good morning welcome back to the explicit measures podcast with Tommy Seth and Seth and Mike app doing some weird stuff didn’t quite show the video intro right so we’ll see what happens we’ll see how if everything turns out okay all right hope everything’s working smoothly good morning everyone good morning gentlemen and a happy Tuesday to you I was a little worried that I was gonna have to run with this run I was like Hey Oh you only saw yourself I only saw myself and I’m like all right I guess I guess we’re doing this

1:01 doing this G have to argue with myself if there was if there was ever a point in time when we didn’t think we needed to rely on each other those are the times that’s the moment oh it’s just me H hi I doing guys ready for the next 58 [Laughter] minutes oh it’s funny no really big announcements I guess right now or today I guess there’s a couple like you

1:32 I guess there’s a couple like you I guess there’s a couple like pieces across the Microsoft blog I know pieces across the Microsoft blog I know pieces across the Microsoft blog a lot of the big announcements I mean a lot of the big announcements I think have already occurred they occurred during the fabric conference that’s when Microsoft announced the many main things we don’t really have any major news announcement I haven’t seen any too too many big blogs or that came out here recently so with that we we’ll just roll into our main topic today we’re just going to talk about where do we see fabric where we see the interaction of interaction of AI and what does the future of Consulting kind what does the future of Consulting look like in lie of these other I

2:04 of look like in lie of these other I guess new technologies that are coming to place inside Fabric and in our PB experiences what’s going on there and how does that got to affect maybe if you’re consulting or you’re doing consulting around those things how does that work for you and I I guess I want to be clear here too this is not an episode just for Consultants right I episode just for Consultants right you in some way are a consultant to mean you in some way are a consultant to other teams or potentially inter interorganizational you’re you’re helping other departments so there is a yes you’re not getting paid like an hourly wage to get that stuff done but in some cases you are

2:35 stuff done but in some cases you are Consulting in your own a company to help other teams or other people out so I just want to add some context there how I’m thinking about the word Consulting Tommy any other any other thoughts on how you want to frame the the topic for today no I I think that was a big point because I was going to say if you are a central bi team the only difference is you don’t Bill hourly and you have a lot less email addresses compared to being a consultant there’s obviously a few other things but no this is a mail bag question that came into US yes a lot more mailbags we’re getting from you

3:05 more mailbags we’re getting from you guys still not a lot of names so we again we want to attribute all these great questions to you guys is the name even on the form is that something are we screwing up we may need that though too so okay yeah maybe that’s what it is know how to do sh kind that way people just start doing anonymous anonymous anonymous anonymous but I think I think this pains very well though in terms of where we’re at in the age and the era and where we’re moving with fabric I and where we’re moving with fabric you’re we’re shifting so fast with mean you’re we’re shifting so fast with fabric we’re shifting so fast with AI

3:35 fabric we’re shifting so fast with AI I’ll pause there because I think it’s time for NPR Seth to come out and and at least phrase the question with the mail so we we have our mailbag voice mailbag voice please great mail bag voice yeah but to to reiterate onto Tommy’s point right this is a mailbag so this this topic comes from an audience member and and we were happy to pick it up because it is very Brant so it’s a short one for all you mailbag listeners out there that love these

4:06 listeners out there that love these conversations this one is not several paragraphs long so you don’t have to deal with my voice forther hi there I’m a bi consultant sorry hi there I’m a bi consultant I have a question with Microsoft Fabric and the recent developments in AI what is the future of powerbi Consulting and services and that’s where we jump off it’s a heavy into the into The Ether before

4:37 heavy into the into The Ether before before we dive into like hey like what is the future of Consulting what why do businesses need Consultants why do why do businesses use consultants and you’re asking the question is a in a pure Consulting kind question is a in a pure Consulting space right not like a like of space right not like a like an internal resource that’s helping other team you’re talking about pure about yeah if we’re talking about the future of Consulting like part of that is why do like current state why do businesses use

5:07 businesses use Consultants so I think there’s two reasons and I just watched a video from Ike Ellis who is also on YouTube but he’s actually out of San Diego area he doesn’t have a lot of views or hits on his videos but his stuff is very relevant and he just did a whole pricing and Consultants spiel about what’s going on like how does work how you should price things as a consultant and what’s a good deal what’s not a good deal I like his perspective on things but it feels like there’s two in my

5:39 but it feels like there’s two in my experience there’s like two main initiatives that require Consultants there are skills that we know we don’t have and we need to go hire someone who has them that is external to the company so I’m borrowing a skill set to get something done or accomplished yep and then I would hope with that engagement the the skills of the consultant are somehow getting worked into the team that they’re working with right producing some documentation working with with them some education so

6:09 working with with them some education so the goal I think is hire the knowledge bring it in use that as a baseline to help educate internal people I think it’s a great way to go and I think the second one is you’re out of people there’s there’s an there’s an initiative that is important enough that you need to get something done and you just don’t have enough bodies to get the work complete so you’re like well I’ll just I know I don’t need someone forever because if you if you knew you going to be able to do a project and you’re going to need someone for a long period of time you’d

6:39 someone for a long period of time you’d hire them right if you had the budget you’d say look we’re going to need this person here they’re going to support for a long time but if you’re doing a we just need some effort right now to get something started we can maintain it we just need the starting point is hard so we need someone to come in who’s already done it let’s buy what they have or buy the the time the effort to produce the thing that’s how I see it I don’t know any other options no I I think for for me and without being salesy at all I think people are looking for a hyperfocused war chest because you

7:11 hyperfocused war chest because you have a consultant who like at least every consultant I know in the bi space or we can also translate that to MVPs they’re going to spend multiple hours doing learning more than they need to and they’re working in so many other areas so from a client coming in they’re like hey we need someone who’s either been here before has spent more hours than humanly possible and and just really I think loves the tool and I think that’s what usually fine with the consultant is like oh I can pick from here this is your situation well I’ve

7:41 here this is your situation well I’ve been in this space or I’ve seen this situation multiple times so let’s take a look and evaluate the people that you have the situation that you’re in let me grab from my War chest and say Yep this is a good this is a good fit let’s try to implement this this doesn’t work and I think you just get those people who are on the extremes again I think in our space especially I I think you say that for all three of us but my gosh the the amount of knowledge and time that we spent Mike this is something that you

8:12 spent Mike this is something that you told me told me is yes all the stuff around powerbi there’s no college degree you don’t go to McKenzie for powerbi corre but it’s not the hours that someone actually spends and the amount of time that they’ve devoted both from The Experience knowledge and skill that really makes up I think where the consultant comes into play what do you think you asked a question I’m curious like yeah so Tommy and I are Consultants so we have maybe a bit more of a unique you were consultant

8:42 bit more of a unique you were consultant previously you did that for number of years have the experience experience consultant yes but right now you’re at the lens of I’m in a company who may want to hire Consultants or you choosing to hire internally so did we did we hit the use cases on how you would perceive why you would want to hire a consultant I I think you nailed them all right like specialized skills you’re you’re coming in there’s there’s an extra boost of productivity or a major initiative but at the same time like the extra boost is also where

9:14 time like the extra boost is also where I would I I don’t look at Consultants the same way I look at contractors right to me contractors have a skill set but those are people I plug into already like driven initiatives within the team we just need the extra boost yeah Consultants I look for for the other two which is specialized skills or kind which is specialized skills or thought leadership Direction because of thought leadership Direction because we’re looking and then that dovetails into like what Tommy said which is more of that outside perspective right like

9:45 of that outside perspective right like we know a lot of things about our company but hey we want to know we want to know what are the best ways to implement fabric right or yeah the next next gen of things where do we need to go does that spool up an R& D project or the net level of things and that’s where I think top tier Consultants or Consultants that are engaged deeply all the time in continued learning and you can well I should say I can tell but it

10:16 can well I should say I can tell but it like it’s it’s hard I think there’s that distinction where you try to make try to sus that out and understand like what is the experience level of the of the people coming in and giving outside

10:25 people coming in and giving outside perspective because the outside perspective sometimes is inval especially when you work in the same place all the time or whatever so couple Industries part of that you’re like oh yeah I’ve seen this before it’s very similar to this other industry it seems to be doing something similar those kind to be doing something similar those things so to to take that though I of things so to to take that though I would say that my experience it like I am a I represent the people who aren’t

10:55 am a I represent the people who aren’t Consultants right now but do Consulting in a business I still okay my experiences as a consultant have forever changed how I will ever work as an FTE because I still I still act like a consultant yeah at least in my mind right you have to deal with obviously there’s a lot more political Games part of an organization but at the same time it doesn’t change how how you can see impacts being made

11:27 how how you can see impacts being made by your work as as a consultant in Oran and the that the perspectives are fantastic because you get a lot of bre of oh this company same technology stack very similar slightly different but like there’s repeated patterns across companies and I think that’s a very eye-opening experience because it it lets you so peeking underneath the covers here for everybody that has never consulted

11:58 for everybody that has never consulted right like your exper your things in a company are very similar to what like challenges in other companies and everybody like level each company is at different levels some are further along on the journey some are a little bit further behind the perspective gives you like oh I can see this linear like where you’re at I’m going to integ engage here but no that like it’s pretty pretty standard it’s like it’s the same stuff in in many different places that’s a good point and at least from the present

12:29 good point and at least from the present and then we can move forward is the transition for me to go from FTE to the the old gig wasn’t as challenging as I thought I think dealing with clients because being a centralized team and always living in that space I always living in that space you’re dealing with fires you’re mean you’re dealing with fires you’re dealing with the the pressures from each department who have their own goals and constantly you have to think about the other situations how does this work for the people here so like I said the biggest difference for me is I just have a lot more email addresses that I

13:00 have a lot more email addresses that I have to handle yeah and and I would say like one of the other challenges of FTE versus Consulting is you as a consultant typically you’re hyperfocused on the net not net new but like the project it’s the thing yes and when you’re an FTE it’s it’s very likely the thing and maintain all the other stuff that’s going on yes there’s always the added responsibility of operation of

13:30 responsibility of operation of you developed and created yeah and that that becomes a challenge with slowing down and why you need an extra boost sometimes I think it’s a very valid point and I think also I’ll add to that is there is a lot of political headwinds potentially on your projects that come with companies in general sure and so I think as a consultant especially being an outside entity right you can say things to the company that they may not want to hear but you can say I’ve observed and then you can like land some Truth moments on here like hey I’ve observed

14:01 moments on here like hey I’ve observed we don’t have a really strong Center of Excellence this this I have seen in other companies that don’t add value and potentially will distract you from things and and sometimes people pick up on it and and run with it and yeah that’s a good point least we should try and Implement that one thing I will also note when when I spent some time as an employee and I was the one or my company was the one not my company but the company I worked for was hiring in consultants and I got so annoyed because I was like these Consultants come in and they ask ask us all the questions the employees the people that are doing the

14:31 employees the people that are doing the work they ask us all the questions and I regurgitate like here’s what I think we should do this is a great way of going this is a problem this is where we have friction I’d give them all the I would tell them the Consultants here’s what’s going on the consult the Consultants would then go away repackage everything up into like this PowerPoint slide deck and then regurgitate literally what we said back to the company and the leadership is yeah that’s a good point yeah we should do that oh yeah that’s a big struggle we’ve always had and then all of a sudden things start happening because the Consultants came in and said it and myling well do what this

15:02 it and myling well do what this reminds me of it it is what it is but like it to me there’s this like so I’ll let me just finish my thought here and I’ll I’ll give it back to you the thought here is they’re saying the same thing your existing employees already have said it but because you’re paying them a lot more money to come in and say something like this or they present it in a slightly different way or they’re like well my other experience has told me and then what I observe from you is this and and you hear this like oh yeah so like sometimes it feels like you

15:32 so like sometimes it feels like you don’t always need Consultants listen to your people so anyways that’s another side toll side conversation you gonna say something Seth it it it’s yes I am and it’s probably going to get me in trouble with that’s okay is this is this reminiscent guys of of anything that that you is a husband say oh boy I’m not that no it’s it’s the it’s the it’s the H honey honey you look absolutely gorgeous today you’re just saying that

16:04 gorgeous today you’re just saying that because you’re my husband like what I what I like it doesn’t matter as much right as opposed to and now don’t get me wrong like these are niceties and things that my wife and I have had these conversations where it’s like no it means something Seth but when somebody else says it and I think like to me I’m chuckling because this is where hey what you’re an FTE but your opinion it matters but maybe not as much as consultant Mike Mike’s talking about the

16:35 consultant Mike Mike’s talking about the ultimate consultants get away with the just saying in the professional world you do you do and for example if I said Mike I really don’t like what you said today I’m just saying all of a sudden it’s like well someone else must have said it too that’s what Consultants can do too they go listen your your Coe is terrible just saying and then all of a sudden it’s okay so we get a the consultants get away with a lot so yes and no yes and no think I think you do like I’ve also seen occasions where the Consultants want the client to be happy and so like again

17:05 client to be happy and so like again you’re here to solve a problem you’re here to make something go like you’re here to deliver something right so I’ve also had the observation of like the bait and switch you get you get the salesy really smart knows everything person coming in the lead architect and then project starts and like okay well now we need to take the lead architect and then need to go find the next job and so you get swapped out with people that are coming in that are not as knowledgeable or overviewing content but like it’s I wouldn’t say bait and

17:35 but like it’s I wouldn’t say bait and switch because it depends are you getting the outside perspective and the extra boost if that’s what you wanted or do you need the outside perspective and specialized skills only that’s true right and sometimes it’s a mix of specialized skills knowing the clearest path forward to set the architecture to set the path to set the road map and then you have the extra boost behind your team right true like technically yeah it’s I think there’s a tradeoff but if we’re going to switch if can we switch gears we’re going to switch gears

18:06 switch gears we’re going to switch gears from like business need right let’s talk about like future State because that’s where the the question was I just wanted to establish because I’m going to come back to it probably what is the current needs of business and then where do we end with that but how how are Fabric and Rec recent AI advancements poised to transform the powerbi ecosystem because I think that’s where more this question is focused like how is the future of of bi or powerbi Consulting and services going to going to change well I think one thing I I I feel

18:37 change well I think one thing I I I feel like there’s a this is just a general observation I think a lot of companies right now are hearing the buzzwords of AI and trying to figure out where they can put in some AI I like I me I was every single random program you sign up for there’s some level of AI attached to it somewhere everyone’s adding something that’s a bit more magical to their program and they’re using some level of either it’s smart math or some interesting thing but they’re they’re calling it Ai and jumping on that we jumped on the bandwagon tips

19:09 that we jumped on the bandwagon tips plus has AI generation for wireframes so upload your images and then we use ai ai to like figure out where the visual should go on the pages like so we’ve decided that that’s that’s a good thing because it it’s supposed to shorten your time to delivery of stuff whatever that thing is right and so I feel like there is this need from companies to say look we’re working on AI we think we need to do more things with our data we’re ready to POS we want to position ourselves to leverage AI

19:39 to position ourselves to leverage AI with our data to add more value to our company and our customers I think that’s the high level goal in that statement there’s nothing wrong with that statement I think what’s wrong with it is I think people don’t gr or understand the size of effort to get clean data to make AI work and so I think a lot I think a lot of companies are more foundational in we’re still trying to figure out where to grab and put all the data first so that we could even give it to someone who could actually go do more of the machine learning and the AI and

20:09 of the machine learning and the AI and the whatever the heck else you want to do I think that’s very similar to the story that’s happening with fabric too there’s so much of a focus on the technology we need to adopt the technology we need to adopt the technology but people don’t have the grass on the skills or the story of how their data is Flowing I want to phrase at least for myself the AI conversation because this is really broad right because when you say Ai and organization you could talk about the development the coding llms you can talk about the

20:39 coding llms you can talk about the prompting that the systems I really see the AI for the consultant and AI as the service for business intelligence in the same way

20:48 business intelligence in the same way that if you had vend diagrams for everything in bi there’s a vend diagram between business intelligence data visualization business intelligence data engineering business intellig powerbi there’s these vend diagrams of the specialized skill and if I were to take the business intelligence AI I’m actually going to take less of the development and the coding and what’s co-pilot and I’m going to think I’m going more around how does the prompt engineering how does the flow of being

21:19 engineering how does the flow of being able to ask questions not just about your data but for the business intelligence team to implement easier ways to get scoping Discovery calls and to answer questions about their data around the organization that’s the skill set that I see organizations maybe they don’t know they need yet but I think that’s going to be where that overlap is between Ai and bi and Mike I think you hit it on the head people are looking at the technology the technology the technology

21:49 technology the technology the technology but what does AI run on how many times we said that it runs on good data clean hopes and dreams is what it runs on right now look at look at every Giant llm model on on hugging face what what’s the other side besides the model it’s the database how large that database is and I I want to jump in here really quick you said a word hugging face I think a lot of our audience doesn’t know what hugging face is and actually that sounds awkward when you you say it out of context of course I’ve been on it too many times it’s normal now so Huggy face. com you can download utilize any

22:21 face. com you can download utilize any model text to text image to text llms it’s a huge space for open-source models like ithub for AI so to speak yes and it’s Huggy face. is the website and it’s it’s like a collection of so I believe this is I believe this is a spot where like code lives potentially so there’s code to help you train a model and then there’s code that actually has like a do a model or some and they have like a lot of different things like you want to do things with computer vision or natural language processing or

22:52 vision or natural language processing or something with audio this this is a website where all the these these models or trained pieces of code have been publicly set down it’s a collection of a lot of things so that way you have a central place to like search for try stuff we used a little bit of hugging face for some of the AI generation for visuals on pages we started here a little bit looking for awes would things here be able to be used so yeah and the last I’ll go and then I’ll get your guys at least

23:22 and then I’ll get your guys at least thoughts on it is I think for the AI bi convergence it’s very much going to be how can I create that GPT or that custom interface where someone can ask questions about from the bi team or get some questions about their data is going to be the skill set that’s going to be required or requested so I think I think this is good though but I’m not sure where so I’m trying to think of like the the intersection of Consulting and AI based

23:52 intersection of Consulting and AI based things and so I when when I think of AI I’m thinking of companies implementing AI to help them produce something more like one area that I think is a really great place that AI is really winning right now is like Creative Image production right you’re doing a marketing blur but you’re going to be selling something right you can provide a prompt to the AI and say I’m going to sell something like this and provide me an image that looks like that I think there’s a lot of

24:23 like that I think there’s a lot of really interesting things that that are coming out from that as well I think those areas are very relevant so if you’re a consultant in those spaces it would be it would be who of you like to learn those Technologies because if someone’s coming to you and asking you hey I need a marketing campaign well great you’ve already got a prompt that helps you build a majority of that that you can run that through and that saves you time to produce the output that you need to go give back to the customer so that is where I would see a portion of consulting stuff come

24:53 see a portion of consulting stuff come from but I there’s other places too where I think mainly where my feelings go with AI are where can companies Implement things with AI where is it adding a lot of time and and and helping their employees I feel like right now code based things are where AI excels from an internal company perspective I’m I’m writing things in vs code I’m I’m writing applications I’m writing Dax I’m doing M I feel like those are great

25:23 I feel like those are great opportunities for AI to support the ex existing team whether or not your consultant uses that or not I don’t know but I think that’s internally that’s how I’m thinking about it but there’s probably other places where AI can be placed so you’re saying more not so much how does the consultant as a service do AI but the consultant as utilizing AI for their services is that what you’re saying because I hear that and I I would agree I would tend to agree with that I think part of it is being equipped to use that

25:54 part of it is being equipped to use that so to your point Tommy like I think that’s a part of it I think in general people should learn how to use Ai and there’s there’s areas where I think it makes sense where I’m also potentially going with this is do we really need Consultants now and can the internals of a company Implement AI in a way that they don’t need to hire that external marketing firm anymore they can build their own prompting and learn it and internally now they can prompt what they need to to get AI to produce what they want so I think I think there’s a lot of

26:25 want so I think I think there’s a lot of there’s a lot of opportunity there for companies to start not needing Consultants but instead implementing AI solutions that would help them keep those costs lower because an AI is going to cost you substantially less than a consultant to show up your most expensive asset in a company I think are the people primarily especially in digital companies and there might be some like Capital expenses for some businesses but the people you typically are your most expensive thing that you’re hiring into a company so in in terms of the the

26:57 company so in in terms of the the question I had like like how are Fabric and AI poised to to transform bi I think we are hitting them all and delving into what AI currently is is assisting us with right it right now and and one that’s missing one I think that is missing that fabric is doing that I see Microsoft doing that others are not is consolidating data right because in order for all of these AI things to work you need the large data Consolidated in an area where you can start to make

27:27 an area where you can start to make sense of it and I think that they’re doing that one like like Delta tables like they’re they’re streamlining the storage of analytical data where we are seeing advancements right now are faster development right we’re getting some Speedy tools right like describe them however you want shortcuts to a lot of manual work that we’ve had to do just code in general right no longer am I going to Google all the time for like oh what does that function do oh like

27:58 oh what does that function do oh like there are co-pilots right there are things that help us build code faster to do the things we want to do with data and I think that’s ultimately like where we’re seeing AI be implemented but if you guys think about what is what is the buzzword mean for the lay business person like for the rest of the audience that aren’t as geeky as us right like what is

28:28 geeky as us right like what is AI for business intelligence in the future what does that look like like not not looking at like the stepping stones that we’re in right now but like I I think where this question drives me is what is that future look like and I get there by saying well what does the lay person think about Ai and AI in business intelligence meaning if you could refine your question because I think there’s two ways to take that there there’s an ideal

29:00 ways to take that there there’s an ideal if I’m a business user right I all I look at is the artifacts that are like currently take powerbi all I all I interact with is the report what does the future look like for me or what is going to be pitched to me right that yeah that we would need to get working I think the lay person’s expectation of where this is going is that ability to ask questions and get the information quickly on what need there’s going to be that in a sense

29:30 there’s going to be that in a sense circumventing the the report or clicking in on three things just to say hey what are my sales can you tell me what I’m expecting to get who are my best accounts or who may be my best accounts right so some of the technology already exists with machine learning but it’s going to be that ability to interact and the ability to ask easier questions with your information than is pre than is currently currently available I I think for the business you it’s it’s going to be one of these areas of like it’s it’s supplementing your

30:04 of like it’s it’s supplementing your existing tooling to support faster things right for example like I can’t stand making slides everyone knows I don’t really like making slide presentations I do for my church I build a bunch of slides and sometimes I I present inside our our user group our live group inside Church powerbi my power my PowerPoint slides that I do for like it’s literally like white pages with words on it like so unting to look at however however like I can put the content on the slides I would rather like have the

30:35 the slides I would rather like have the AI come through and say I’m going to make this whole thing pretty I’m going to stylize all the things that you’ve done so that it’s like white text on a colored background or white text with some interesting things on the like to me there’s a lot of remedial things that I feel like I’m doing that could be just done better with some AI to like give me a lots of selections of things that I can pick from and then produce the output so you said what is business know you said what is business intelligence an AI sit Seth I think the question is very good if you give me my crystal ball and say where do we put this as a very large task what does this look like many

31:06 large task what does this look like many years down the road I think you start talking to the AI to have it you

31:12 talking to the AI to have it you basically bring it piles of data you add relationships to said data and the model should be able to build itself there should be things in here that are just very common like again sqlbi has done a great job there’s a lot of Dax patterns that occur right I should be able to talk to an AI I should be able to have every numerical like if I get the columns defined right these columns are numbers these columns are not I should be able to talk to an AI and the AI should be able to spit out 30 measures here’s all the basic measures sum of

31:43 here’s all the basic measures sum of average of do it should happen like instantly and you should be able to use that it should be able to go again they’re doing some things right now at Microsoft with parbi right now where you’re saying ask it a question it’s able to Define on a page it says here’s the visuals page I can write a summary for you on that page so I would think there would be more questions or general questions that could be asked against the data and then what will happen is the model will build its own relationships between tables it will create its own measures and it will

32:14 will create its own measures and it will then produce tables or visuals to Output information for you and say Here’s here’s like the same way that that chat GPT or does it in Google Bing chat whatever the thing is co-pilot in in bing right you ask it a question here’s what I think is the answer and then it provides like reference notes as to why this is this is true that I think is where we would want things to go are we there yet probably not and I think you’re talk we’re talking about the base technology and as

32:44 talking about the base technology and as I frame this around where would the consultant come in or more importantly where would the organization or that lay person find frustration I think initially with this technology is it’s not going to do exactly what they want and just like we talked talk about bi culture or data culture I think there’s going to be an AI culture I think the consultant’s going to come in this hyperfocus being able to take these toolings that the technology will be available but it’s going to be very basic it’s going to be a foundational technology not everything for your

33:15 technology not everything for your organization how can we get these prompts how can we get these Technologies already available to wait for what the organization is expecting because if I’m asking for sales there’s really no technology that’s going to do that without some customization around that organization’s data and I think we’re going to find oh yeah look the technolog is available but no one knows how to engineer that or no one knows how to customize that so it just works so to

33:45 to customize that so it just works so to speak that’s where I see the consultant coming in where yeah we’re going to work with your data but we’re also going to work with these whatever technology that you have to ask questions and we’re going to make sure it’s waited for the sales Department the marketing department which data are you looking for what’s important to you because we know none of that technology just works yeah I think I think one of the other things that the the future potentially holds that we haven’t talked about is how I I think there’s a real

34:18 about is how I I think there’s a real possibility that AI starts making decisions on its own right like we put it through batteries of the these models these things I want you these models these things I want better more accurate predictive know better more accurate predictive forecasts right I want these things and to to drive into AI making the simple decisions because we’ve talked about like the tooling itself right allowing us to make like build the things faster but I think eventually it’s like okay yeah Mike to

34:49 eventually it’s like okay yeah Mike to your point build the things faster but I also like here’s the outcome I want I want you to I want you to decide between these two paths or three or four or 5 and give me the best one or give me the output of the top 10 best ones and and that’s very interesting because that is going to require a whole lot of work to ensure that the the model or the AI is producing the expected outputs

35:19 the AI is producing the expected outputs and why right and that’s where I think a lot of a lot more analysis and understanding from individuals is going to be absolutely required but at the same time like if I look look at the complexity of business intelligence and and I say we we we talk about this all the time we get yelled at because we’re we’re finicky about roles right like what are what do what do we do as bi developers you cross the spectrum of data engineering all the way

35:51 spectrum of data engineering all the way up to business conversation right that ecosystem of data is still an unknown known artifact in the business by and large like how that all happens so that’s why I asked the question the way I did is what is AI and bi look like for people who have no idea what we do yeah it is the chatbot it’s the thing that I expect that in the future I’m gonna have a chatbot and I can ask it anything I want anything in sales and I and what what comes out of

36:24 sales and I and what what comes out of it has to be the right answer not not sometimes it has to be the right answer and if you think about I guess just the amount of work that it goes in that we still have to manually do but still has to be done on some level to structure data to run it through it analyze it transform it certify it

36:54 analyze it transform it certify it validate with the business that these are the calculations that me these things and we produce a report where we have to justify that now you throw in all of the other ways in which we potentially need to integrate still need to build these things still need to go through all these processes like those pieces aren’t going away it’s just a an automated computer that we train is

37:24 automated computer that we train is going to hopefully in the future and I think it will solve a lot of that for us but it still requires somebody to understand that you’re all of these things are working as we would expect them to so I think is it going to change is the job of a business intelligence professional going to change yeah some of it’s going to be a lot faster a lot easier but at the same time like if I have one trustworthy metric that’s coming coming out of an output regardless of who is asking or how I

37:57 regardless of who is asking or how I can’t have Betty in sales going give me the sum of sales and Frank in marketing going give me the sum of sales and two metrics come out or them not understand that th those are specific differences and it’s agreed upon within the organization like so I when I look at it from the Optics of how does business interact with the insights and the data that we are already taking a ton of time in developing or trying to get to the

38:28 in developing or trying to get to the point where we can produce those insights I I think the same problems exist the only differences more organizations either can get there or get themselves in trouble faster and I I want to hyperfocus on something you mentioned because I think this is the real overlap between both of these call them different Industries but man they’re going to be a lot of overlap you said about show me that predictive sales well in order for people to make that jump to to assume and really rely on

39:00 jump to to assume and really rely on that it’s the same problem we have now with data and it’s trust and I think that’s going to be the biggest barrier to entry when it comes to whether it’s in your data or it’s in that random your own chat box that you have it’s trusting that information that output where you still struggle with this now and I think furthermore the idea with the different sales I think there’s going to be a fur elevation of the semantic model

39:30 be a fur elevation of the semantic model in terms of its importance of a single model because us as a bi worker man we’re gonna have to really focus on which columns and Fields and metrics are weighted in that in that box in that model that’s going to be a huge and I think it’s going to be on the bi team or the bi consultant see I I feel like I hear what you’re saying and Seth I really agree with your point with like when there are important decisions like there’s got to be someone

40:00 decisions like there’s got to be someone looking over what is coming out of the AI and saying is does this make sense are we are we really going to make decisions like if we’re going to hire more people or we’re going to fire more people we need real solid information to like make those really strong business decisions I totally agree with that but on the other hand I’m looking at this going like the AI is just not good enough yet to understand which of my columns are important enough to me to make sure that I’m writing reports like so like to me there’s this whole idea of like how do we educate the AI on what

40:30 like how do we educate the AI on what does it need to have in order for it to be able to produce good insights or good things back for us and so one of them is like look we know that there’s this date table we know that there’s going to be all these like cumulative sum over a year like it it does this weird stuff like oh did that sales are higher when you cumulative suum and and the month number is 12 like it’s yeah we know of course like this is duh but that’s not adding any Val like that’s an Insight that’s not adding any

41:01 that’s an Insight that’s not adding any value to what we’re doing in this AI space so I feel like to your point Tommy right there’s there’s there’s this concept of right now we’re seeing AI do the simple things right AI is creating measure descriptions it’s creating measure explain this measure explain this deck explain this code that’s very simple stuff but if you took that to the next level and said okay if we if we look at a model and said okay if we went through this and let the AI determine okay these

41:31 and let the AI determine okay these columns we think are more relevant you columns we think are more relevant I should be able to say to the AI

41:36 know I should be able to say to the AI what’s the data quality of this table and it should be able to say your your probably your most reliable columns are this this and this and most of your information comes here and you have a lot of blank areas in here right so I think where AI should be supporting us a lot more is we should be able to ask more general questions about individual tables and then again handling some of that that lightweight work of like making sure we have all the simple things done then the AI should be able to come back and say okay now that we’ve approved all the names of desre

42:06 approved all the names of desre descriptions on measures tables and columns the AI should now be able to say look these columns are the one that says sales would have more weight than the one called sales key right so it it should also get smarter in that as well so to your point Tommy the AI needs more labeled data to support these are the things in the model that we think we should be caring about focus on those aspects of the model first and and produce insights based on that information and I I would call that a bi

42:37 information and I I would call that a bi service moving forward that’s just going to be part of the service offering whether it’s a consultant or whether you’re the centralized bi team that’s a good point it’s a tool it’s a tool leveraging different tools yeah they don’t care like ultimately when somebody hires you they don’t care how it’s made unless unless you’re doing the training of like developing developer right training developers or the next gen of selfservice like ultimately it’s the it’s it’s just shifting a tool or leveraging something that’s out there that speeds up the process which is just

43:07 that speeds up the process which is just B more beneficial for the customer oh no in a year from now I’ll put some money down I’m G to be telling clients like hey one of the things we do is we also wait models in some capacity and I think that’s going to be just part of the offering why doesn’t why can’t I offering why doesn’t why can’t the AI the do you’re gonna you’re mean the AI the do you’re gonna you’re gonna see see this is where you just lost me oh oh okay oh you’re my okay so fundamentally this entire time what we’ve been talking about is like profound expertise in the

43:37 about is like profound expertise in the particular areas of all of the data things and in order to vet Ai and the new tools and all the things that are coming it it requires you to know what you’re doing so where you just jumped in my mind is you have all the data science background and experience to know all of the models that are being generated and that’s going to be part of your service no no no no no no no no I don’t think I I said that I think we’re getting there with the technology I think with some

44:07 with the technology I think with some technology be able to weit it because again we’re talking about this fabric powerbi space we’re coming into know how to wait your models which columns are important I so I think it’s things we’re trying to do today okay so you’re talking about the value of of data columns Etc okay very much what mik was talking yeah I see honestly yeah to to cess Point waiting models is a term that’s been around for a while around machine learning I think I’m speaking more of where we’re heading very much

44:39 more of where we’re heading very much okay cool all right cool see so so where regardless of where we go I think what we’ve seen so far is AI is very good around the structured things where do we where are we finding value it’s in code where’s the next steps Mike related to data well it’s going to be in the where we are already giving structure to the AI right like so it can read that because their structure where I think we’re far away from is this pipe Pie in the Sky Vision that many organizations

45:11 the Sky Vision that many organizations because it’s the buzzword automatically think is I can use a chat bot to ask anything and get the right answer out of it and I think that is much further away I agree having said that how crucial is it to stay a top these things I think it’s imperative oh yeah because and the reason for that is we’ve been in this arena for a long time and you’ve seen exponential increases in technology powerbi was a GameChanger right big

45:41 powerbi was a GameChanger right big shift in the market how it was able to do all these things with the underpinning in the structure all of those are starting to formulate and we’re already getting and talking about like the tools that are helping us speed up I do think I am of mind that I think this one is a sharp jump like all it’s going to take is Bam the next level is here Mike and all of the sudden AI does a fantastic job of categorizing and classifying your your columns like

46:12 classifying your your columns like setting all the data types building a model for you etc etc and you have to be up to speed and engaged here to understand when those spikes happen because that’s where I think the the the fastest to accelerate in organizations that are plugged in to wanting to engage in leveraging AI to gain better faster something I think I think there will be those Leaps and Bounds

46:43 will be those Leaps and Bounds and those that are riding that curve are are probably going to reap the most benefit and put themselves out farther ahead than other technology advances before because if you think about it if we’re talking about structuring our data pulling it all together using fabric pulling these leveraging what we have implementing the things that work and our speeding time we’re already familiar with this ecosystem all we have to do is go figure out how to how to plug that new spike in versus companies

47:16 plug that new spike in versus companies that aren’t or individuals that aren’t engaged in the AI conversation and what its offerings are through the different tools will will there will be a point where they’re scrambling they’re to try to figure out like oh my gosh how do I get to the point where everybody’s been that has been riding this this real curve and that’s where like I I am I’m one of those guys that you’re I think you’re the further out on this wave you are versus the way things have been done in the past where a lot of organizations don’t ride that front wave I think R& D

47:48 don’t ride that front wave I think R& D in this area is is absolutely one of the best investments you can to briefly emphasize that Seth I’m fully expecting or putting the skills experience documentation now around by the end of this year I’m either writing a contract or talking with someone about prompt engineering that that word’s going to be coming up in a contract at least I’m setting myself up that way because I I’m I think there’s there’s too much convergence there’s too much overlap between what’s happening now and what we do in business intelligence and that may be somewhat of a segue I think

48:19 that may be somewhat of a segue I think with with fabric and I know we’re we’re running a little low on time but man alive we’re seeing so many other things with fa to where it’s it’s not a I can’t call it a product it’s seven products yes and I think I think this is one of the neat things so we’ve been saying this I think for a while now on the podcast and I would agree very much with your with your comment there Tommy it’s like we are looking at fabric it’s not not just too there’s too many thoughts in my head I’m s I’m sorting out thoughts here

48:50 head I’m s I’m sorting out thoughts here fabric is the collection of all the years of knowledge of Microsoft building other databased data data tools right so pipelines is as your data Factory you pipelines is as your data Factory synaps brought right over so spark know synaps brought right over so spark and synaps and a lot of the other notebook notebook writing SQL SQL Server lless the SQL analytics endpoint a lot of those things are just lifts of other stuff that Microsoft has already built and it’s in the cloud already so to to that point like great now it’s all in one tool where I pay one license fee and I get all the things and

49:21 license fee and I get all the things and I use them how I need to for my business so one it’s hyper flexible but think is great where I think this fits with AI now and how this lives in this in this new context with now fabric is I think we’re going to have business users show up business users are showing up with with their existing skills in power Cy and to I would argue they’re doing data engineering they’re just doing it in Excel in in non repeatable ways or a little bit extra work right what we’re now giving them with fabric or powerbi

49:52 now giving them with fabric or powerbi is a better way to automate some of those data movement things and so we now have the ability of bringing a whole bunch of business users that are being hey I just used to build reports and I used to build very simple models or even complex models but I didn’t have all the tooling to simplify my model and push that Upstream now we have the ability to actually do roach’s Maxum which is transform the data further upstream and produce the tables we actually need to simplify our data models so I think the right path here is look at all the business users

50:23 is look at all the business users showing up to fabric educate them on better data engineering because it because it exists and start with data flows work your way knowledge wise into pipelines then start learning notebooks that to me that’s a great progression because it’s it’s becoming more and more code based as you go down that path then once we have a pool of business users turned data Engineers now we start having the ability of okay now what oh like well we

50:54 ability of okay now what oh like well we already have the tools available to us to start doing AI or predictive things and so now those same business users turn data Engineers can now start doing lightweight data science work now I’m not saying it’s going to happen all out of the out of the gate but I think there’s a progression here of people liking what they do continuing to pushing in more learning more and things because the tooling is not quite good enough yet for data scientists and data Engineers to come over to fabric yet I think it’ll get better give it some time

51:24 think it’ll get better give it some time I think they’ll be more useful use cases in the future and and then we’ll we’ll do a better job of making all the feature Rich things we need to to bring those people in but right now I see it as bring business users in educate train and give them more capability than they ever had before Mike you rais a very heavy question a very interesting question in what you said and I’m I’m intrigued really to get both of your guys’s opinions with fabric if it is seven products seven products that have existed that have been full-time jobs

51:54 existed that have been full-time jobs for a lot of people individually is the expectation whether for the

51:59 is the expectation whether for the consultant or for the FTE I see expert yeah yeah to be an expert in each of these products in fabric is that going to be a requirement or an expectation I think it’s going to be so I’ll mean I think it’s going to be so I’ll say it this I’ll say it this way as a company looking at other people I think when you see the word Fabric and now that there are so many other tools inside there I think the expectation is you’re looking for Consultants who know all the things about all those parts of the fabric I’m not sure we’ve had enough runtime on

52:29 I’m not sure we’ve had enough runtime on fabric and there’s been some new products that have shown up that have not been anywhere else like kql custo they have existed in Azure but maybe your mileage will vary if people have experienced with them yet data activator a net new thing yes it feels like aure event hubs and there is some borrowed principles from other things that are existing but at the end of the day like I I feel like you’re hiring a company and the company needs to have the compl the complete skill set of

53:00 the compl the complete skill set of what’s going on there and then give it some more time the people will actually have more exposure to doing lots of projects across all those different things so I I think it’s it’s not a good assumption for p companies to say I’m going to hire a person and they will know everything I think that’s going to be I think that’s a very misguided assumption I think you may have to hire a company that has multiple people people that know all the things right there’s there’s

53:31 things right there’s there’s hey I’ve got a guy for the data engineering hey I’ve have a person to help you incrementally load this data there’s patterns that we’ve developed over time that help us load these things into our Lakehouse right so I think you need to have again if you’re hiring someone you’re hiring someone who has a team of people that have all the knowledge I think you would need to bring that Tech stack forward but I don’t I think companies are going to continue to assume they can hire the person and they will know all of fabric and I think that’s a misnomer I would agree with you and in the future

54:03 would agree with you and in the future probably disagree and the only reason is so in the same way the expertise is going to be need needed to shape organized data so we think right but even to your point Mike shouldn’t AI be able to model and create measures and do all the thing yes so I think if do you need to be an expert in it all right now absolutely you absolutely need to have an expert that understands how to ingest transform

54:33 that understands how to ingest transform integrate shape data Etc build semantic models all of that still absolutely required in the future no because I think if we talk about the same things as far as like the tooling getting better and better and better and better fabric is an ecosystem that is it hard to imagine that all what were all of these Services all of the sudden now play in a space where a a layish a not technically technical trained user right can create solutions

55:04 trained user right can create solutions from end to end I think that is a very real possibility so do they need to be an expert in all those areas to produce something for the business no do you still need the expert to vet what they build yeah yes like you would want that but but in terms of democratizing data in the in the solutions like in an organization you’re you’re bringing that self-service to people who want to solve these problems on their own and I think eventually is there a world where I believe that this this service and

55:36 believe that this this service and ecosystem provides that I think that the technical capabilities are absolutely there and the paths that we’re already seen to get there are in motion I just don’t know what the future means is that two years is it five years I I don’t know right but at the same time that’s what’s very compelling to me about fabric is if I if I take what I know about Ai and I take like all of my understanding of what I see how it’s

56:06 understanding of what I see how it’s helpful the things that Microsoft is doing they like obviously I’m a Microsoft Fanboy I don’t go study other Solutions out there but like it it is very compelling and I think all of what they building in terms of the building blocks lead me to believe that this is this is going to be one of the Premiere if not the premiere Analytics tool that’s going to be fastest to Market with AI and and all the things and I think that’s the great area kind and I think that’s the great area my last thoughts around Alex is

56:37 of my last thoughts around Alex is saying 12 months Max Alex Powers msft now is he speaking for Microsoft I don’t know know right but put it on the board it’s on it’s in the chat it’s in the chat my my my last thoughts is I think that’s the greatest area Seth is again fabric is not just as much it is seven individual products the big selling point what M Microsoft’s really trying to do is it’s like data engineering for everyone it’s the self-service side

57:08 everyone it’s the self-service side around these tools so maybe the requirement or the skills not so much having that data engineering expert just come in and fix everything because maybe it’s more around helping each individual Department set up their own ecosystem because that’s what Microsoft’s really trying to sell it’s not just we have synapse in a different user interface it’s we have all of our data in a centralized place for each department I think to me Seth that’s the grayest area that I I see where where does the consultant fit or where does

57:39 does the consultant fit or where does that specialized skill sit for an organization around fabric yeah and I think that remains to be seen as things shift specifically for my closing comment to this this question about what the future of business intelligence Consulting is I think it it it’s wide open i i as long as you’re sticking to the advancements in AI I I don’t believe it’s going to strictly be like the more AI becomes prevalent the more business intelligence roles

58:09 business intelligence roles completely just go into Data engineering I don’t believe that because you still need the skills to work with and understand what the business needs are so you’re vetting the outputs as well so are you doing a lot of engineering and integration work probably but in different ways like it I I think it’s just the tool sets that we work with that are going to shift and our priorities could potentially even sit further with the business to say hey man it only took me a week to build this whole thing now let’s spend a week and really dial this in as opposed to hey it

58:41 really dial this in as opposed to hey it took me a week and four days to build this whole thing and now I have a day to sit with you and make sure it’s dialed in right the the the refinement of really great outputs and analytical insights I think is where eventually you could spend a lot more time but yeah In terms of where I think the mean in terms of where I think the future of bi Consulting is I think it’s bright I would agree with that one as well I still think Christian made a comment in the chat a little while ago that said he sees it being in two fold areas there is new users showing up

59:12 areas there is new users showing up needing a little bit of help AI can AI can support that and then there’s those medium and in in advanced users where there’s a lot of repetitive tasks that are going to be have to be done over and over and over again it can help with that too and I think right now where I see the landscape of AI and internal consulting or external Consulting like I it’s not going to replace them it’s definitely not going to like take away that that knowledge skill however I think it will supplement a lot of the the needs there and it would be wise for consultants and companies to

59:44 wise for consultants and companies to understand what is AI how how can I leverage this stuff for what we’re trying to produce in our company and it will shorten the time from I have a problem and I need to deliver a solution I think we’re going to continue seeing that that timeline from statement of problem to completion of solution continue to shrink further and further and further as more and more AI becomes more and more useful and either does the pay I’m solving those simple little problems over and over again or it’s

60:14 problems over and over again or it’s actually providing bigger analysis efforts and and shortening that up as well I think your point there at the end Seth was really spot on I agree I think I think that’s where this is going to go so I’m I’m very like I like what’s going on I think it’s going to be great I I’m excited to see where Microsoft takes us and how it gets developed there’s probably a lot more things to think about here that we’re not even touching on yet but again we don’t know we don’t know so we’ll see where AI is going to continue to fall on the spectrum of Consulting Mike I very like

60:44 spectrum of Consulting Mike I very like it too so well probably they’re gonna just take all of our podcast funnel it through an AI algorithm just that’ll just be the output that’s going to be the eth product by fa we should make the explicit measures AI to an analyze your internal company policies and process so that would be interesting anyways with that all said I appreciate everyone’s time here thank you so much for listening this has been a great discussion also the chat has been very Lively today thank you so much chat for jumping in we had a ton of people on chat just having great comments so

61:15 chat just having great comments so thank you so much appreciate all that and our only ask is if you like this content if this felt interesting engaging you like the conversation that was happening here please share it with somebody else let somebody else know you like this video we’d love it if you would share this video and put your comments below hey I found this interesting because of XYZ things or this is ringing true for me or I totally disagree with this like I this is how I see it we’d love to have more feedback across the community around what you think about our topics and what’s going on in your head related to

61:46 what’s going on in your head related to particularly AI Tommy where else can you find that podcast you can find us in apple Spotify or wherever you get your podcast make sure to subscribe and leave a rating it helps us out a ton do you have a question an idea or a topic that you want us to talk about in a future episode like we did today head over to powerbi. com

62:20 thank you all so much and we’ll see you next

Thank You

Thanks for listening! If you enjoyed the episode, please subscribe and leave a review—it helps others find the show.