PowerBI.tips

Subscriptions as a Silent Hero - Ep. 516 - Power BI tips

April 3, 2026 By Mike Carlo , Tommy Puglia
Subscriptions as a Silent Hero - Ep. 516 - Power BI tips

In Episode 516 of Explicit Measures, Mike Carlo and Tommy Puglia unpack the latest Power BI and Microsoft Fabric topics from the show. You’ll get a quick read on the episode’s biggest ideas, why they matter, and where to dig deeper in the full conversation.

News & Announcements

  • No linked announcements were available in the episode description for this post.

Main Discussion

This episode covers the major themes, opinions, and practical lessons Mike and Tommy surfaced during the conversation. The transcript below captures the full verbatim discussion if you want the exact phrasing and context.

  • Mike and Tommy react to the episode’s biggest Power BI and Fabric developments and explain what stood out to them.
  • They connect product announcements to day-to-day practitioner decisions instead of treating the news as abstract roadmap chatter.
  • The conversation highlights where teams can move quickly, where they should slow down, and what tradeoffs deserve attention.
  • They share candid perspective from real project work, which gives the discussion more practical value than a headline recap alone.
  • The episode mixes tactical advice, opinionated takes, and a few forward-looking predictions about what listeners should watch next.

Looking Forward

If this episode’s topics affect your current Power BI or Fabric plans, use the transcript and linked resources to identify one concrete change you can test with your team this week.

Episode Transcript

0:03 Explicit measures. [music] Pump it up. Be it high. Tommy and Mike lighting up the sky. Dance to the day to laugh in [music] the mix. Fabric and A. I get your fix. Explicit measures. Drop the beat now. [music] Pumpkins feel the crowd. Explicit measures. [music] Explicit measures. Drop it loud. Good morning and welcome back to the Explicit Measures podcast with Tommy and Mike. Good morning, Tommy. How you doing? doing? Good morning, Mike. Good to see your

0:33 Good morning, Mike. Good to see your face. face. It’s good to be back here again. So, we had took a little bit of hiatus. We were doing some recorded episodes. This has been like two weeks or so. We haven’t really chitchated together yet. So, coming back together here for recording another episode. So, this is a pre-recorded episode, but we’re jumping to things this episode. Okay. Well, people know the real time, but this is after the Microsoft Fabric Conference. confence. We got about a week out here., and we’ve we’ve kind week out here., and we’ve we’ve a lot of things that have been of a lot of things that have been announced at the at the fabric conference., we have a whole episode

1:04 conference., we have a whole episode coming to just dedicate and talk about fabric conference. It’ll be next week. So, don’t worry, we’ll get into it. We’ll get into all the details as well and unpack what was released and all the impacts it’s going to have on us as businesses and how that’s going to work., but I I just want to maybe touch on our main topic and I’ll come back to one. So the main topic today is going to be around subscriptions is our silent hero. So subscriptions. Oh boy, this will be a top more subscriptions. This this will be a topic I think. Okay,

1:35 This this will be a topic I think. Okay, before we get into that I do want to give a general feeling for while attending. So this year there was the Microsoft fabric conference that was the a lot of the public facing was in Atlanta. We had a record number of people and while at the conference they kept noting this is the second largest conference other than Ignite which was 10, 000 people. So there this conference was 8, 000 people deep. Wow. Wow. Super big.

2:05 Super big. from a data platform conferences we had the past we had the very original ones data insight days. It was like 2, 000 people initially, right? right? And the whole vibe was just different. There was a lot going on. It was very busy. yeah, it felt like a really big conference. Really liked it. I thought it was great. Yeah. Even BIS summit which got became merged with the business applications wasn’t that large. No. Yes. No. Yes. They went from the the data insight summit to the business application

2:36 summit to the business application summit. summit. Yep. And that was not just dedicated to PowerBI at the time. It was PowerBI, Power Apps, Dynamics. Yeah. And that only had basically a dynamic conference. Yeah. With a little PowerBI on the side with a little sule, little soule of Yeah. And I think I think Tommy when I when you look at like the spectrum of like where PowerBI fits, it fits in this data engineering data platform space. Well, while PowerBI is a part of the story, the broader story is data stuff.

3:08 story, the broader story is data stuff., all all the things that go with it. So, I think fabric and having PowerBI become part of that family. This is the first place where it actually feels like it fits. Yeah. Yeah. Properly. Properly. And and and honestly, this makes sense. Well, actually, I’ll say two things. I’m surprised at the number of people, but it does make sense for it to be its own now dedicated conference because it is six careers into one. I’m going to forever to say that until that changes, which I don’t know if it’s going to, but it it is six careers that you’ve merged

3:38 it it is six careers that you’ve merged into one platform. So, yeah, you’re going to need it. That being said, to be the second largest Microsoft conference out out says a lot. Yeah. I don’t know if they’re going to increase the size at all, but the conference center is huge. It’s really big. big. Yeah. Yeah. Oh, anyways, let me go back here to like my main thought here, right? So my main thought around this one has been over the last two years when fabric was released I feel like the conferences we went to Tommy they were exciting there were some announcements but a lot of the announcements that were made around the

4:08 announcements that were made around the conferences were like we’re adding capability we’re bringing more CI/CD it felt like a lot of the announcements were just more like we’re fixing things right it’s not quite we don’t we we’ve heard you we’re listening we’re trying to fix stuff so this conference felt like the first time that we were moving away from we need to fix more things for you to we actually have new features. We actually have integration with this new thing called agents and AI and developing with AI and agents to help

4:39 developing with AI and agents to help you create. There’s a lot more language around like the creator AI helping AIs help you generate or create new things. that really excited me and that that this feels like a turning point for me where we’ve gone away from like just getting it off the ground to new features, new excitement. I was really excited leaving the conference this year. I was like thrilled. I was like this this feels right. This was like we’re building finally the correct

5:10 we’re building finally the correct things. Microsoft is closing a lot of the gaps that I felt like we we have in the AI space. It just felt like they were like not listening or not getting there. I don’t know really know what it was but just felt like we were behind and this is the first time I’m feeling like okay they are listening they are pivoting we’re moving more towards this agent agentic building space with with fabric so that was extremely exciting for me for me or other conferences have been we’re going to show you a feature but it may or may not come out from six months from now and that was the other problem too

5:40 now and that was the other problem too like hey this seems really cool when’s that coming out we don’t know but this actually real things coming out and one thing I love Mike about you and I we have the passion we both have for the agentic side. We also come from two unique perspectives where you very much come from and look at it from the individual the developers AI tool their tool belt their arsenal and not that you don’t focus on the other side but for me I always look at the lens of okay how is this going to be applied in a business

6:10 this going to be applied in a business and I feel like both of those were answered to to a point at the conference where both sides of for me as the developer which is so important what AI tools am I going to have at my disposal to help build? But then also what AI products am I going to help provide solutions for, which I think are two two sides of the coin that we have to consciously be thinking about. There was a couple notes in the keynote that talked about like the GitHub CLI and like how the GitHub CLI is is helping you like

6:43 CLI is is helping you like and it was it was the GitHub copilot CLI which was what they were talking to that that was the agentic platform that they were integrating into some of the pieces of fabric. I really like that because GitHub Copilot is by far one of Microsoft’s best products around around building things with agents and using agents to help you create things and you agents to help you create things and it doesn’t really be it know it doesn’t really be it there was a word here that was used at the conference a lot which I think is a bit new to people it’s this thing called the harness right so if you think about AI the large

7:14 right so if you think about AI the large language model is just you you sends a prompt to it and what comes back is an answer when you think about the harness the harness is the language around the piece of software that connects to the a like it wraps around the AI the the the large language model and that was another thing that I think I observed from the conference there was a lot more talking about like what is they didn’t say it in every single meeting but there was some conversations around the harness VS code is the harness we’re

7:45 the harness VS code is the harness we’re building other experiences in fabric that are going to use agents and that was like the harness around the agent piece So when you start talking a bit more about how these AI models work, the software around them like anthropic and what they’ve done with cloud code, that’s the harness. VS Code and how you talk to it in VS Code desktop, that’s that’s the harness or the CLI and how you command line interface, how you use that directly, that’s the harness of what it’s going. So there’s a lot more language around that. I think it felt a lot more to me, it felt fresh. Hey,

8:17 lot more to me, it felt fresh. Hey, we’re we’re building these things. These harnesses are going to appear. We’re going to make it easy for you to get access to these AIS. And the last thing I think I would maybe note here from the conference, a vibe that I was feeling was everyone was talking about skills. Well, there was a lot of MCP that up. Yeah. Yeah. Yeah. There’s a lot of MPC MCP talk, which is tooling that the agent can use, but more so than that, every single team was talking about, hey, we’re building we’re working on developing these skills. We’re making skills that can do certain things. And

8:49 can do certain things. And I really like this idea of skills. It seems to be really portable and fast for the team to develop them and leverage them for fabric. So that seemed to be a very common theme is we should definitely be working on and learning how to build our own skills. Seems like a a reasonable expectation right now. This is why I love talking to you because sometimes our minds are completely alike. I was literally going to bring up something about skills and you did it for me and I think just

9:19 you did it for me and I think just important here because honestly Mike I’m finding to your point with co-pilot though they have their own now CL tool just copilot rather than the G8 command line tool which is great. Sure. And the problem though is with skills is they’re hard to import if you have different machines, different users or if you’re using different providers. And I actually built well I built but I have a skill vault. your agent build I think that yeah that my agent build but now what I use that will actually push and sync it will actually do a

9:50 push and sync it will actually do a system links and junctions over to different places because I’m finding a huge a huge influx or a renaiss not a renaissance but this this growth in me building skills yes for different things and what that was another thing that Microsoft really pushed too that not only were they talking about but they backed it up with a ton of skills that they built for the agentic side for Microsoft fabric and I was very pleased to see that that are now in my skill vault

10:21 now in my skill vault and and that’s I think where a lot we’re going to see a lot more improvement around things skills are very lightweight they’re easy to develop they they run on top of so this is where I think a lot of these pieces come together right the harness is the code you use to interact with the the agent skills are something you plug into the harness and so there’s this new world of thinking about how these things get pulled together. I think the general sentiment with people at the conference was AI is here. We

10:51 at the conference was AI is here. We should start figuring out how to learn it. it. And And already I’m looking at like the community. There’s there’s been a couple posts on LinkedIn since the conference talking about people have been like exploring and and diving deeper into to your point Tommy these skills MC MCP servers working on different parts of the project. I’m seeing people generate now full reports. Now, they’re not super polished and clean yet, but people are starting to build like simple reports with Agentic experiences. People are now building semantic models and

11:22 building semantic models and modifying them. Most of what I feel like people are doing right now is they’re using less code and they’re describing more of what they want. I saw a post just yesterday on LinkedIn talking about how a gentleman was taking I think he was taking two semantic models and trying to merge them together because they had two like just two versions of basically the same data or something that was somewhat different and you need to take those two models look I want to pull them together threw it at an agent and said here I’m just going to describe the requirements here’s the requirements I need to take

11:52 here’s the requirements I need to take you these two models figure out what tables are common don’t repeat them right here’s some measures bring these two things together figure it and it did a really good job. He was very pleasantly surprised about how well the agent was able to merge these two models down to one single semantic model. I think this is the start of a lot of other posts and items on the internet around,, Microsoft’s introducing this. People are starting to explore it and now we’re going to see a lot of value add from it because people are going to start again back to the community, Tommy, it it always goes down

12:22 community, Tommy, it it always goes down to to when we’re talking about it. Yeah. When we’re talking about lakehouse or data brick or not data bricks, data warehouse, lakehouse, those things, it’s always like, well, which one’s with the more performant one? Well, the community determines what that is. Experts go out and test it and then say, here’s some of my results. Here’s what’s faster. Oh, I was running this job on,, data warehouse and then notebooks and then data flows gen 2 and let me give you the comparison of the CUS for each one of them. And

12:52 of the CUS for each one of them. And then you can make an educated guess. guess. this agentic world, Tommy. Yeah, go build your stuff in M code, it should be really easy for you to migrate it into notebooks now. Well, I I I envision and I actually already put this in our backlog. I think we need to put a few episodes together just for skills, but I imagine for a lot of BI teams, just like before, your most critical shared resource were templates and themes. I think that most necessary,

13:23 and themes. I think that most necessary, most prized resource is going to be skills and that being shared across an organization or at least a BI team. Sometimes your predictions are a little bit off, Tommy, but I think this one’s going to be a very solid one. I I I [laughter] think I think this one’s this one’s spot on. on. Hey, metric sets. Yes, we know metric sets was a bit off, but yeah,, yeah,, task flows, too. You win some, you lose some, but I Yeah, Yeah, I don’t think I think the concept is really solid. I think to your point Tommy for creators of content in PowerBI like this will be the way you want to

13:53 like this will be the way you want to build. I also think there’s another level here around who can create right previously interesting like on our previous episode of DAX doing DAX in PowerBI right that’s kind doing DAX in PowerBI right that’s where I feel like this is going of where I feel like this is going you had people could like generally build some simple DAX measures and get started down the road the the route here but like when you start doing like conditional formatting I want some SVG things that are maybe measures there’s there’s just parts of the DAX experience like I’m doing some complicated

14:24 like I’m doing some complicated filtering. Yeah, Yeah, there are some other aspects here that just gets really difficult to get DAX to write itself to write DAX correctly. You have to really know what’s going on. I think the addition of the agents are going to make this easier. And I think what you’re going to start seeing is not just remember how time we had like the analyst and then like the data modeler, right? I think what this is going to do, it’s going to give more capability to the analysts to build better DAXs on their own or more DAXs on their own. And

14:54 their own or more DAXs on their own. And so we’re probably going to get a proliferation on more DAX being created by more business users, but they’re actually not writing it. They’re talking to their agent about what they want. I have a what your take for you then. I’m going to propose something for you. See what you think right off of what you’re saying. Okay. And then we’ll we’ll get into the topic. Sure. If I am a team that has a shared set of BI skills, agentic skills that were built for the team, creating DAX, modeling, etc. Can I am I are you going to be

15:25 Can I am I are you going to be comfortable allowing junior analysts, maybe someone who started a month after in the company to start building reports or modifying reports if they’re using that shared skills, they already have access to the skills. Say, can you update this? But there’s someone who’s very new to the company. That’s a great question and I’ll even double down on your question, Tommy. I went to a session that addressed exactly this topic. Interesting. Okay. The session was AI versus novice or

15:57 The session was AI versus novice or something like that. something something along those like and the the concept here was look I have an expert in the room who’s going to build a report andor some measures and then I’m going to give the same report or the same starting data model to a novice and say okay now you build something so the whole the whole session was voling back and forth between like okay here’s what the expert built here’s what the novice built can you determine what’s real what which part of this the expert build versus the novice we had some tax measures we had

16:27 novice we had some tax measures we had some four pages. We kept volumeing back and forth and the audience was trying to vote on like which area was made by professional or not. A great session. It was a good session. Very good session. It was fun. It was we laughed a little bit. It was it was just interesting really at the end though. So they they did this really fun like voling back and forth and some of the things that the agent did was very good. Other times it just was a little bit off for the novice. They didn’t quite know what to do. And so the the end of the session spoke to okay look if you give

16:59 spoke to okay look if you give AI to someone who’s new it adds a lot of capability but they don’t know how to guide the AI specifically to always get the exact right answers. Yes they can build a lot more. Yes they can build a lot of things inside the model. they might overbuild the model a bit like too many measures. the the the agent will just is just trying to make you happy. It’s just trying to fulfill your request. So if you say hey agent make any measures in the model it may make 30 measures all of them may or may not be needed but it’s writing a

17:30 or may not be needed but it’s writing a lot of things. So on one hand it’s it’s helpful that way. in the creative the UI space the UI side it didn’t really do a great job of making the UI it was okay really having a user involved in the UI creation was I think the better mode. However, However, when you look at the perspective of novice and expert, if you give AI to a novice person, they do a little bit better. They get a little bit better quality, they they build better stuff.

18:00 quality, they they build better stuff. But when you give AI to an expert, you get like this 10x effect, right? So the message here was the experts took a lot more time to build visuals, report pages and all the things to make like the measures performant and work well. But when you gave them the AI, instead of having them take two or four or eight, like there was towards the end when we’re doing like the visual building with the AI, the expert took like 40 plus hours to build something

18:31 like 40 plus hours to build something with the with by themselves, like just clicking and and doing things while the the novice with the AI was able to build it in like 10 minutes, 20 minutes. It was like it was like substantially different of like the amount of time taken to build something. And so I think that’s the story here is it’s AI is a multiplier on what skills are already there. It’s not going to necessarily give you like expert knowledge on everything. It’s going to you still need to be knowledgeable about the topic

19:02 need to be knowledgeable about the topic or domain, but the amount of time it takes you to build it is going to be way shorter, right? right? And and that’s the power of it. So, you want to get your experts up and running on agents and building with agents because it’s going to make them much more effective and go faster and that will give them more time to be able to think about what testing should be doing. Does this look right? Let me focus more on the requirement side. And I imagine that’s also your experience too because that sounds very much like my experience where I’ll give you

19:32 my experience where I’ll give you just a very quick workflow here where a new project came in new data one of the things is off the transcripts that’s where I get a lot of my tech specs the discovery that [clears throat] was able to build a lot but there’s a lot of I’m using the MCP and I’m using a lot of the agentic stuff to help build the model but that doesn’t mean I’m done like I’m going through afterwards and then you have to verify like Okay, how did it build it? And as it’s telling me what it did, it’s like, no, we don’t need that. No, we need to make sure there’s relationship

20:02 need to make sure there’s relationship here. So, there’s still that expertise. There is the fine comb fine comb brush that the expert has to do even if AI can still build something. It doesn’t mean it’s in the in a place where I’m going to send that to production. Yes. Yeah. It’s it’s it’s going to help you build at scale, right? Right? But you’re you’re still going to need to be able to like reason about what it created, right? You still need to check it, right? So,

20:32 right? So, a novice won’t know how to tell the the AI know, right? That’s the point here. here. This is that’s a great point, right? That’s that’s that’s the weakness of giving AI to people who are not as familiar. However, I will say I think this I think the knowledge of this whole thing should be okay. The agent will build things. How do I know it built it right? So, whatever you’re building, you need to think about talking to the agent and say, I want you to build these measures and here’s how I need you to test them.

21:04 and here’s how I need you to test them. Give me the results back so I can verify that that that the agent built something correctly. And I think that’s really the lititness test. Yes. of is it actually building what you think you thought it should be building. So I think if you add that little bit of flavor to it, I think it actually adds more capability across the spectrum for both novice and experts, you’ll get better results out of the AI regardless. Yeah. Okay. Awesome. Awesome. We’ve talked enough about like random things at the beginning of this. This is the news and articles. There’ll be more topics. Trust us, they’ll be coming for FabCon. We’ll we’ll unpack

21:35 coming for FabCon. We’ll we’ll unpack FabCon. I think we’re going to talk a lot about all the new features coming out and the big announcements that Microsoft made., one of them being the data database hub, I think is what they’re calling it. Make sure I get that quite right. I think it’s a good story. I like it. We’ll see where this goes. And that being said, let’s jump into the episode. Tom, you want to give us the the main topic today? Yeah, this is going to be one of our last mailbacks for a bit. Again, we got to cover FabCon, but thanks everyone again for their feedback and also

22:05 again for their feedback and also sending in some great questions. So, here we go. Just some feedback comments on your episode about subscriptions. While I’m sympathetic to Mike’s defense of the PowerBI platform as a way to draw people in to analysis and research, I am empathetic. Wow. Sympathetic. Empathetic. Nice. When it comes to getting actionable insights into the hands of professionals who may not always need to be digging into PowerBI reports, here’s a use case where the email subscription has proven useful. We

22:36 email subscription has proven useful. We have a sales tool that contains the name of professionals who you would want to

22:39 of professionals who you would want to contact to advance a sale. Periodically, those people leave the company, but the data source that records that departure is not directly accessible in the sales tool. I built a simple little report that compares the data in the sales tool with the data source with departure information. If names in the sales tool pop up as departures, the report shows a big red number counting the name number number of departures. There’s about a dozen different people who is responsible for

23:09 different people who is responsible for updating the sales tool if one of their contact changes. I set up an automated email subscription that sends this report every Monday morning showing the count of contacts who knew to be updated or none if there are no departures. The email the email shows a page with the metric and a list of departed names if any plus the link to the report. The report itself contains a link and instructions if you need to update a name. Dead easy. You are you see either a red number or none and

23:39 a red number or none and whether you need to look deeper. This one little tool probably saves hours each week and is reassuring reminder that our contacts are clean each week. And that proposes to us us in defying our face on how we currently feel about email subscriptions, Mike. And I think a really really powerful argument in favor of email subscriptions.

24:13 It is. I Tommy, as much as I like rant and rave about like we have to get away from email, it’s antiquated. I don’t like this is just my vibe in general with with emails. Emails are slow. It’s It just fills up inboxes and I don’t like using them. I get so many emails nowadays. It’s just difficult to like I don’t it just feels antiquated in this new world of agentic everythings emails just feel like dinosaurs to me in

24:43 just feel like dinosaurs to me in general right and and so I understand this concept I’ve pushed so hard against this topic and like we really don’t want this there’s other ways there’s better ways of doing this it keeps coming up it there’s just some organizations I think Tommy I think I’m just going to just you Tommy I think I’m just going to just bite the bullet here and know bite the bullet here and Some organizations just love email. That’s how they operate or that’s the tools they’ve been given to communicate to each other. It is what it is. You’re going to have to have some of these things emailed out to people. I’m going to make an argument here

25:13 I’m going to make an argument here against what you said here. I don’t think it’s every organization or some organizations feel that way with email. In the case of this mailbag that we had, there’s a very clear actionable insight of what email can do and more importantly what the report can do. If you were to think, Mike, what you would love to have every Monday to tell you what to do, regardless if it came as a PowerBI report or just an HTML email, I would take that. If I said, “Hey, here’s

25:44 would take that. If I said, “Hey, here’s all the S. s I haven’t sent out yet or here’s what’s still in the pipeline who I should contact. I would absolutely love that. In fact, Mike, I’ll do you one better. Gmail has has a new feature like that. It’s called ATC. Outlook obviously has been really trying to build on this with Copilot. To me, when you look at email subscriptions, you and I have a negative taste in our mouth because we’ve seen all the poor examples of what email subscriptions are. Yep. Usually a lot of people are just sending out a normal report as an email

26:16 sending out a normal report as an email subscription. That is a not only a bad case, but I think that it’s a negative it brings a negative taste in people’s mouth, a negative experience of what email subscription is. This example here is a perfect example and really maybe a more of a hot take is one of the few examples where email subscriptions really shines. I have one clear action I need to do. There’s not 10 different things I need to look at. It’s not a generalized report. The point of the

26:46 generalized report. The point of the report has now been built into be more of an application than a report. Yes. Yes. And I think that’s really where here email subscription shines. But you have to think about it. The intention of the report was built for email subscriptions. I I think that was the that that feels to me like the intent of like pageionated reports or SSRS reports., that’s Yeah, that’s where I’m going to go back. But I guess I guess my my rub on this one is

27:18 I guess I guess my my rub on this one is you just rubbed me the wrong way with that comment. But well, I I think I think the rub on this this comment here is just it feels like whenever you send something to email, that’s the end of the action list and you’re setting up process for people to try and get them to do something else above and beyond that. Like, so, hey Tommy, I’ll send you the email. [snorts] It shows up in your inbox, but what’s really the action that you need to do? Let me give you let me give you an example of how I like I

27:48 give you an example of how I like I would like to use email, but like maybe slightly differently, right? right? If this So, for this email here, right, this action, right? Sure. Sure.,, I’m going to try and pick on some other items here, right? It contains names of professionals that you would contact to advance a sale. Okay. This is a list of names. So, I’m going to just land something in your inbox. Probably that’s already where the work is happening. So, here’s the here’s the inbox. Here’s the report. Here’s the list of names you should go contact. [snorts] [snorts] I have some internal processes already

28:18 I have some internal processes already in my business. We sell some software. I have a contact form. I have people fill out things. I’ve been building new agentic workflows that take the form information, one determines if it’s a real person or not, then two stores the data. Then three researches the customers that that user’s email like the domain name of it. [snorts] Are you a real business? Go to the website, research about them. And then three, come back with a here’s a

28:48 then three, come back with a here’s a proposed email title and subject line that I would want to give to that user. So I’m actually so in my process I’m not only doing the cleaning and like you only doing the cleaning and like verification of the data I’m also know verification of the data I’m also building like the next step. So, okay, okay, I’m I’m trying to think about what is where is the human in the loop of this? Where [snorts] So, I would look at this and start saying instead of having it send you an email with just a list of names,

29:18 names, why not build a system that grabs the data, pulls the list of names, looks at their history, does some reasoning with this? cuz this is the part that’s like now for that list of names I’ve got to go through and like talk to every customer or have a template set up and have it have the AI reason through this. Hyper customize each message to the customer. Hey, we’ve got sales. We haven’t seen any sales from you recently. We’ve got some new products out. Here’s some announcements you may want to be interested in. Like,

29:48 you may want to be interested in. Like, but that should to me that should land in my draft inbox. Like, not even in my inbox. It should just it should just build me stuff and put it in my drafts. And that way, every morning I show up, I go look at the drafts. Okay, overnight here’s what the agent decided it thinks it should do. And either I either delete the draft or I modify the draft and send it to the customer. So to me, I I see what this is doing here. I’m looking at what we can build now today and feeling like this I would want to

30:19 and feeling like this I would want to evolve this process slightly and not rely on people to do the process because in in I I think people would be more happy to do it if you gave them a bit more recommendation or suggestions or starting point to get them further down that path. I miss disagreeing with you so much. All right, few things here, Mike. You’re talking about Yeah. You’re here. You’re talking about an ideal world where we already have a lot of things in place, especially from the agentic and

30:50 place, especially from the agentic and the AI side adopted, which as we know is less than 5% of companies right now who have successfully integrated or adopted AI. So, right off the bat, Mike, while that may be a good case for Mike Carlo and a smaller organization to be able to get up and running, if you’re working in a larger organization, first off, again, you’re making the assumption that AI or anything’s already integrated, adopted, text adopted this, we already have the data in the right format and structure

31:20 data in the right format and structure to do so to get that implemented. So, that’s the first I think barrier to your point. Now, all this being said, I agree with you. There can be an evolution here, but when you talk about the here and now with a lot of organizations, if you were to propose what you said, maybe there’s a few people who would agree with you, but if I’m a sales team or a sales department of 200 or more, that’s going to be hard to get up and running, especially from the AI side. The other side is really

31:50 the AI side. The other side is really focused, hyperfocused on a sales department. Mike, I don’t know if you ever when you were FTE, if you actually were part of a sales department, but there’s also a behavior thing that you have to go, you have to overcome to get past that, especially from the evolution. You have to meet them where they’re at. You may call that less evolution. You may call that,, they need to get on top of it. But there’s a there’s the game that, you there’s a there’s the game that,, that we have to play here. And for know, that we have to play here. And for me, I know when I actually I’ve been

32:20 me, I know when I actually I’ve been able to work with two sales departments when I was as a a BI person part of the sales team. And one thing that I learned very quickly was you may not have people who are the most technologically advanced or want to be, but they’re good at what they do. And we met them where they’re at. For them, they live in email. A lot of sales department, especially B2B, live in email. So this is a great use case where I’m already checking my email. I can already see the contacts because this is

32:50 already see the contacts because this is where I’m already living most of my life. It’s that and my phone system. Those are the two places where they’re living. They are contacting their contact records be it dynamics or Salesforce and email is where they’re spending 90% of their day from a technology point of view. to give them extra is another adoption curve that again not saying it can’t work but I think you’re to me you’re speaking of something to an ideal state where there’s already too many barriers today

33:20 there’s already too many barriers today for that to become a a viable solution as of when we’re talking today in 2026. Oh man, Tommy this is interesting. So I think this topic pulls on a piece of information that Satia So I saw I saw a short a YouTube short Satia was being interviewed by somebody and in that interview Satia was describing where does does the business problem solving or the

33:52 the business problem solving or the business logic or the business context live right now. today currently. Yeah. Yeah. Today, right now, a lot of our business

34:00 Today, right now, a lot of our business context lives in apps, applications that have been built, software as a service, those things, those those things, right? And and to some degree, right, that’s baked into fabric and PowerBI, right? That’s where the business logic is being invested or built into to produce these results, right? Hey, I need an email sent out. I need to have a list of of people to adjust. But Sat made the point that he said in the future, right, what we’re seeing right now is a rapid move of that business logic out of the apps and into

34:32 business logic out of the apps and into agents, agentic experiences, a, the large language models. That’s that’s skills. Skills are that thing, right? Skills are are the are the compiled knowledge. And so when I look at what business are trying to do, right, let’s let’s step back and look at this problem holistically, right? If we just move back and say, okay, what’s really happening here? Really, what’s happening here is we need some automation. We have we have some ideas around

35:02 We have we have some ideas around context, someone leaves the company. If the customer doesn’t have a related connection point to our business, we have a problem. That’s the problem. If you just step back and say the problem is people move on. We need to make sure every customer is handled well and has a contact person. That’s that’s the problem of the story. If we don’t think about PowerBI as part of the solution and step back and say let’s let let’s give our agents the experience,

35:32 let’s give our agents the experience, this could actually manifest itself in a number of different solutions, right? This this could be a little mini application. this could be something that automates it and and I think to me when I look at this problem I’m going okay this works dead easy I get it you built the process it works but is there a better way to integrate more more forethought into this and again I have to be very clear Tommy what your team uses for tooling you you have to build

36:02 uses for tooling you you have to build around that right if if email and teams or if email is the only thing you use and there’s a dynamic system that you only use and that’s what you give your sales team, you have to work in the boundary of that stuff, right? That that’s the that is the container by which you must play. But I really feel like if when what is happening now for me is I’m able to step back a lot more and say let’s just identify the problem and then let’s just think creatively of how we solve that. Maybe there is a

36:34 how we solve that. Maybe there is a salesperson portal that does this. Can I even go further down this route of like,, not even in this example here, can I do something that doesn’t require you to step in at all? Can I have it auto do a lot of like how much further can I automate this? Can I take the problem and actually solve it without having any any person in the loop at all? This may not be a good example for that, but that’s where I I’m I’m fundamentally rethinking how to solve problems now. Does that make sense, Tommy?

37:04 Does that make sense, Tommy? No. And I so I think this is a good point where as a consultant coming in to have a few questions on really on what the constraints are because that’s what you’re talking about is I’m coming in I’m saying they want a better process. We have to first work on again what playground are we playing? How big’s our sandbox? Yeah. What are our constraints? And a lot of organizations to your point they may say what we would actually love we love this PowerBI email subscription thing but we’re open to other ideas

37:35 thing but we’re open to other ideas we’re open to things agentic we actually want to have people spend time or less time in apps less time in applications and if that’s the case then to your point there’s a lot more we can build so I agree with you coming in especially more as consultant where we can have that that persuasive power to be able to do that. But again, the human element here, I think, is so fundamental where we don’t want to also disrupt too much either.

38:06 want to also disrupt too much either. And that’s the other side of the coin. I can provide,, build it out of an app, put it in their email draft to your point or have another system where there’s a contact thing where things are already built out. Yeah. But again, everything around technology is always going to rapidly change the process. People process technology. the pillars we’ve been talking about for 400 episodes now plus. plus. I would I like your idea there, Tommy. I think rapidly change the process. Yes.

38:37 think rapidly change the process. Yes. But I also think we need to rethink the process and figure out if there’s more that I could automate now. So I think one of the piece that was hard to do in software was reasoning. How can you make how can you make the software make an educated guess as to what the user is doing? I think the large language models do a much better job of like reasoning through things, giving it context and saying here here’s the information about this customer synthesize for me an email to go get them engaged. I’m seeing a lot of things right now where people are building like search engine optimized solutions and websites and hey have AI

39:08 solutions and websites and hey have AI run all your marketing branding stuff like you can outvolume your customers. Like there’s a there’s a lot of interesting things here, Tommy, that I think are very useful when it comes to AI things. But again, I’m still trying to sus out like how do you people are finding value from this? How does this going to change my actual workflow? What does this actually do dayto-day for the teams that use Outlook and PowerBI? Like what what does that look like? Yeah. Let me give you another hot take or a little hot take here too. Sure. What if you looked at a whole

39:38 Sure. What if you looked at a whole organization and all their processes? This is not one I would attack first. This is ripe. I think Oh, interesting. Here’s what I’m hearing you say. Here’s what Yeah. Here’s what I’m hearing you say. You’re looking at each process, especially the mailbags that come and you’re looking at it in the lens of how can we in a sense identify this? What are the different ways? Is this ripe for something we can transform? However, I look at this and I say, I bet you there’s four other things in this company at minimum that we can look at

40:10 company at minimum that we can look at where Agentic Solutions would be much more powerful. This by itself works. This does not seem like a terrible crutch on the company the way it’s built. Could it be identified? Sure. But does it need to be? And I this is where I want to challenge you a bit where have to be identified because I guarantee you there’s a lot of other processes that I could look at this company. Maybe it’s getting writing their contract. There’s a great example where again we only have

40:40 a great example where again we only have so much resources to do things and this is probably not one I would focus on. I think to me this is looks like a great solution. Yeah, I I agree with you Tommy. There’s probably other things, right? That’s not the topic that was given to us., there’s there’s a million things you can automate away and make it more efficient, but, given that I’m just given this point to the question. Sure. Sure. Right. I can’t really speak to the hypotheticals of like where else you can apply

41:11 apply more automation to what you’re doing. I So, let me ask you, let me ask it a different way. Does this need to be identified the way it is? It could be, but does it need to be? The lemon may not be worth the squeeze on this one. I think you to your point, you could probably add some different process here. Maybe where I would challenge things a little bit further would be, does it have to be an email? Right. Right.

41:41 Right. Right., let me let me just pull back some some thoughts here. Right. Mhm. I think this is a solution where we cater to the tools that we’re in. I have a lot of reports now that I run and build that I go to the report that that’s where the work happens in the report. report. Yeah. Yeah. I think what we’re seeing here, Tommy, if I had to like really distill like let’s I’m trying to take a a good lens on this, but then also distilling like what’s going on. There’s this need that

42:14 what’s going on. There’s this need that we’re seeing here where there’s an intersection between a report providing data and you taking some action and doing something on top of it. So one item here as well would be okay we’re sending emails there’s a couple pages here but this also could be served very well with translitical task flows. So inside the report, okay, here’s the report. Here’s the list of names. You show up, here’s your list, right? For each one of those names, you

42:45 right? For each one of those names, you could stay inside PowerBI the report and continue to do things and continue to edit data. So one of the one of the weaknesses that we’ve had traditionally, Tommy, has been like we don’t have the ability in the report to like basically edit or modify data easily. And now with translitical task flows with DAX sorry userdefined functions in fabric with with SQL databases or in fabric on SQL all these other experiences now make it actually a little bit easier for us

43:16 it actually a little bit easier for us to change data in line directly in the process and have it do things. So So I would look at this process and say is there any way I could keep people into in in their workflow longer? Do we really want people to stay in that report? So,, I’m hearing some things here. Look, here’s your Monday report. This is,, every Monday the team needs to do some activity. Okay, fine. It’s only once a week activity. Again, to your point, Tommy, earlier. This is lightweight, right? This is

43:47 This is lightweight, right? This is probably like five minutes of anyone’s time. They open the email, it’s part of their workflow. Do I need to take any action? No action needed. Move on. Delete. You just delete it. Move on kind Delete. You just delete it. Move on thing, right? That’s probably fine. of thing, right? That’s probably fine. Where I struggle is this seems highly catered to like a very small running process. This I feel like could be quickly growing into other things where everyone’s getting daily emails of data, tables of data sent to them every day and then they have to download it and then work on it separately. So I think I think there’s

44:17 separately. So I think I think there’s there’s to me there’s a little bit of a difference between this is a very small use case, use case, not run very frequently. needs action on a beginning of the week from everyone on the team, right? This I think fits very well for like the email subscriptions. So, I’m not I’m not going to poo poo the idea. I do think we need to continue reviewing these kinds of processes because there potentially is other future things that are coming that will make this easier for us to build with. I like Yeah. You see what I’m saying? Like, no, no, no. I like where you’re going

44:47 no, no, no. I like where you’re going with this. But then you begin to touch on you begin to touch on another part where we may not have the ability to influence influence and I I’ll that’s true. So then this is a big part. So I actually share a lot of sentiment with the person who sent the mail back because one of my first jobs one let’s be real. Let’s be here. You don’t have sentiment. You wrote this. This is Tom. [laughter] Youly I’m going to get Mike on this one. I’m going to write my I’m going to write this email. I put my name on it. This is Tommy’s [laughter] question back to us.

45:18 Tommy’s [laughter] question back to us. Yeah. So Mike, but Mike D. Carlo sent a message about AI and Tommy sent a message [laughter] about email subscriptions. But no, I I share this because I remember when I when I started the company, when I moved to Chicago, they were like, “Hey, we have this thing we get with Excel. We want the same thing.” I said, “No, we could build a PowerBI report.” We built the report, but they’re like, “Can we get this subscription?” Yeah. And there’s this frustrating thing on but you have the dashboard and Mike it wasn’t until we started sending these few people and it

45:49 started sending these few people and it wasn’t a lot of people sending those few people the PowerBI report as an email subscription was did we see the adoption take off now the more important thing is what you said that I think was important here if we wanted to do anything different outside of that that required approval of a change of process that I did not have especially as a business intelligence expert. I would have to change that sales team’s and again as small as it was y was y I would have to change their process

46:19 I would have to change their process which I did not have the ability to do. Correct. Correct. Even if I wanted to change technology and I think that’s one thing you’re speaking to here. If I want to change change the technology and the platform that they use, that’s going to require a change of process that most times even as a consultant we don’t necessarily have the ability or the authority to do. That’s a problem and that’s going to be a problem with this agentic side. That’s going to be a problem getting people off of email where we don’t have that

46:50 of email where we don’t have that ability. we can propose that but if they’re like no you we’re not going to change everyone’s date everyone’s Monday okay well now you have that constraints that we have to live in and email subscription becomes really powerful yeah this is I [clears throat] think you’re bringing up a really solid point here Tommy is like when you start building reports that span multiple teams and multiple different processes you have to operate in the world that they live in right right right and this example here that team doesn’t have the ability to log into a custom app or is it it’s again reading

47:23 custom app or is it it’s again reading between the lines here for them. You’re right. Yeah. Reading between the lines here, they have access to PowerBI and they have access to email. Like that’s that’s the extent of what we have. So if we can’t play in that world, we can’t make it we can’t improve the solution in any way. Right. Yeah. I agree with you. Good point. Right. I agree with you. Like that has to happen. Yes. Yes. where I I don’t I’m not this is not a push back but this is where I think as B as BI people so I’m so I’m going to

47:54 as B as BI people so I’m so I’m going to change my conversation here right I’m going to change my conversation up to like leadership of teams that are allowing this to happen right you’re solving problems here right leadership should take note of this this and it’s going to be up to us as the developers of this space to continue to inform leadership of hey we’re doing this saying, “Here’s something that works. We’re grabbing we’re gaining some adoption, but if we added these, like we need to know the tools so well that we could add additional features to it that

48:25 could add additional features to it that improve the process that make it more automated.” And I think I’m going to go back to like my earlier concepts, Tommy. When we start really pushing into automation of things, when we really start pushing into like how efficient we can be, that’s when things start shifting a bit more for me. Right. Right. I love what you’re saying here, but continue. continue. But that’s my message to leadership is here’s what we’ve solved today. This is a progression. It’s we can’t get there

48:55 a progression. It’s we can’t get there all at once. And we’re not going to just throw away everything we’ve done before and say, “All right, no one uses email at all ever. We’re not going to like that’s not going to happen. That’s not the that’s not the world we live in. So, what are we doing regularly to progress things moving forward? How does that look?” And I I think that’s where I’m I’m focusing more of my my thoughts and attention now is okay, where’s the next step for us? Do this, make it work. Does this ever get improved? If not, if there’s other bigger fish to fry, go work on those things. But leadership needs to be thinking about

49:26 needs to be thinking about AI, AI, PowerBI, this space is going to continue to evolve. It’s going to continue to change. change. So, how do we as leaders push forward? How do we as builders of these systems push leadership inform leadership about there’s better ways of doing things? Maybe we maybe we should have integrated like at some point hey is there a better way of do can we put this back in dynamics like if that’s what you use is there something can I embed a report there that is where

49:57 embed a report there that is where people are again it’s just trying to how can I can I so many thoughts trying to distill them [laughter] how can I simplify and distill what we’re working on so that people have less friction that that’s usually what I’m going after I I hear what you’re saying here who needs coffee with Good statements like that. You’re getting me riled up here. I’m very excited about this because you I’m very excited about this because, but here I think this is what bit know, but here I think this is what bit what you’re trying to say too. You want to be a leader in your organization. I think you want to take the next step in

50:27 think you want to take the next step in your own career only talk about technology if asked because what you’re telling me here, what I’m hearing is when I start taking a leadership role, whether as a consultant or in FTE in my space in fabric in the agentic world, world, I’m working on towards the solution. What are you guys trying to improve on? I will give you that solution. Oh, you want to know where the technology is? I’ll tell you, but that’s not the goal of our conversation here. What are we trying to do here? And I think a lot Yeah. Yes. I like this. This is what I really like.

50:58 I like this. This is what I really like. So, take the technology out of it. Like, what are we trying to do? And then And then what what I believe is happening here is is the technology under the hood here. It works like this email subscription may work for now. Great. No problem. Right. Do they even have access to any anything agentic? Maybe not. So, we stay here. We’re done. Leave it. It adds value. Move on. Great use case, right? Stay with it. However, I feel like over time, we’re continually going to move the ball. We’re going to continually

51:28 the ball. We’re going to continually shifting things. Stuff is going to continue to be evolving into these more robust systems. The technology is going to continue to shift. Fabric is going to get more agentic. And what I feel like we’re starting to see is fabric has made PowerBI less reporting and more transactional in nature. It’s adding more like features to it. And I think that’s the part where I’m like okay like this comment and question

51:59 this comment and question let’s revisit this one in six months right does this change in lie of announcements and progressions that Microsoft has made on top of the product I would argue the answer might be maybe. Maybe this is a good time to revisit this. this. I could argue it would change now if you get the right question to you. If someone comes to you says, “We have this current system, but we want to improve on it.” And your response is, “We’re going to change it to an application. We’re going to do something agentic.” You already lost because you’re talking about technology.

52:29 about technology. I would argue so much and I do this all the time and this is what I’ve seen is the biggest change for me and I think especially in what you’re in the light that you’re talking about Mike is is when we start coming with what are you really trying to achieve here I get this current technology that you have but the process the goal of the team here what are we really trying to achieve I guarantee you if they come with certain situations all of a sudden you can change the process

52:59 change the process But if you come,, with air guns and arrows of here’s how we’re going to change the technology, you’re not going to change anyone’s opinion. opinion. But if you’re going to change with how we’re going to save time and maybe maybe their goal, Mike, to your point, is not just to check,, who left, but it’s we would really like to send out faster,, people who left and get those people on board and on a call. Okay. Well, that gives us the ability to figure out a solution. And again only talk about the technology if asked.

53:29 talk about the technology if asked. There may be something aentic behind that but our goal more than ever in this space is not about just knowing the technology. Yes. Because again you can throw an AI problem or find an AI solution and try to find a problem. It’s not going to work that way. I think this is really I think what you’re addressing here is really the crux of the issue, right? this I think too often we focus in and we narrow in on the technology piece

54:00 narrow in on the technology piece and that’s a miss because the technology should just be the mechanism by how we get business done the means to an end a means it’s it’s it is the means to an end in reality we should step back and say okay that’s not not really what we want let’s let’s just focus on what is the problem and how do I get to the solution And that may evolve over time. I think what Satia said earlier when I was talking about earlier Satia saying like the business logic is moving from

54:30 like the business logic is moving from the apps and software into the agents and how those things are doing. Who who’s to say this doesn’t in the future turn into like a skill. Yeah. Yeah. That just automatically does this. Hey, I want this stuff done. I want to and and really where I’m going a lot of this is I think while this is a good report, this is a good analysis here for me. I like to think about how do how does this how can I automate more? How can I suggest more to the user? Right? I like workflows where it does some research,

55:02 workflows where it does some research, adds things I can’t go do or would have to go do initially to and and then make educated guesses as to what I want it to accomplish. And then I step in and say, let me review your output, right? Does this look good? Hit one button, approve, and move on. So, anything I can do to reduce my mental load on that that and automate more of it, that’s the place to spend our time. And again, this may not be the process., , and and I do think like this is a great use case for pageention reports like

55:33 use case for pageention reports like right. right. Yeah. Yeah. Getting people’s attention. All right. Don’t fit that in. Don’t try to fill that in. Oh, pageant reports work. All right. Let’s end the show. No, [laughter] well, we we’ll take it where it is. I well, we we’ll take it where it is., I I dislike I dislike mean, I I dislike I dislike Are you telling me you’d rather have pageentionated reports than email subscriptions? If I’m email subscribing things, I think I’m doing pageentionated stuff. That’s where I feel like I’m sitting. I I just I I just I don’t I don’t love having like a

56:04 I don’t I don’t love having like a PowerBI report with like a big like PDF or image thing. Like it doesn’t make sense to me. Like it’s it’s totally uninteractive. But again, if if you had the normal again, if you built a report to be a report and then you start email subscription. Yeah, I agree with you. But again, this was a report built for email subscription. There’s an intention here. I agree with you. True. Yeah. And again, pageion you again, every pageionator report I I think we’ve gotten in the mailbag came from Mike D. Carlo.

56:34 from Mike D. Carlo. The the one the one note here maybe that would be interesting and and again, I

56:38 would be interesting and and again, I have not tested this. I do not know if this is accurate or not. So this would be something we’d have to like legit test, right? If you had built a report to run, again, this is probably like at volume at scale. Does the report getting sent as email cost more than the pageionate report getting sent via email? I don’t know the answer to this one. This is purely speculation and the difference may be so minimal that it’s like not a problem, not an issue, right? So we may it may be doing nothing. let me just ask one note

57:08 nothing. let me just ask one note here too, Tommy. Yeah., Yeah., does a PowerBI report allow you to when you schedule it? This is just me lack of like I don’t do this. I try and stay away from this because I just don’t know. Can you adjust dynamic filters on a PowerBI report and have it when you’re doing the emailing? Okay. And that also I know that for sure works in the pageant report space because that’s kind pageant report space because that’s the idea, right? Hey, render this of the idea, right? Hey, render this report with Real Security for Tommy. That way only the things that Tommy cares about is rendered. Yeah.

57:39 cares about is rendered. Yeah. It basically loads the report, renders the filters onto the report, prints the report, and then sends it to the email. Yeah. Yeah. Yes. Yes. That that was relatively new feature. I think in the last four years that came out. So, okay. So, given that, right, that makes the two tools very equal as well. I just I just feel like I don’t want to everyone all the time to go do stuff. Well, I’ll I’ll give you a m a micro point of a closing thought here because I think this serves well

58:10 because I think this serves well in terms of our big theme here about the technology and where does it fit. So when I the skills that I built actually around building so now project planning for for work initially there were some tools that I was trying to use a few years ago that it just did not fit with the way my mind worked. I love notion but I would try to use it for work but trying to write everything in I became disorganized. Now that there’s some agentic side to that where it notion can anytime I’m

58:42 that where it notion can anytime I’m doing a meeting it says hey do you want to do a transcript? Well I can connect that to a project. It has these database relationship between pages. Then there’s these new custom agents in notion where I basically say hey look at the transcript. Let’s build out what the deliverables are on the project. Well, the fact that now these te this technology is feeding into other ways. My process has changed, Mike, and because of the technology that I’m using. So, the way I’m handling projects, the way I’m handling my

59:12 projects, the way I’m handling my daytoday has drastically changed because of the technology that I’m using. And it it works out so much better that way because my process is now whole bent on the technology that I’m using because it now finally fits. It did not work that other way. As much as I wanted my process to change before because I wanted to use a certain technology, it just didn’t work. Now that the techn is at a point where it really makes sense, I’ve changed my

59:43 really makes sense, I’ve changed my daytoday because the technology finally fits in into a workstream which has now evolved my how I approach my projects, how I make sure that things are done, my getting Yeah. done list and I think that in a bigger scale is how you look at other organizations and look at departments. I I like that concept, Tommy, because at the end of the day, we’re not trying to use technology to solve problems, right? Problems exist and we’re trying to use technology to solve them, right? right? And so I we have to continue to evaluate

60:15 And so I we have to continue to evaluate and keep moving forward. I’ll also argue to this question today. Look, if you find a solution that works, run with it. Let it go. And the only thing I would maybe provide some guidance here around this one would be is every so often just take a look. Does this get do we need to improve this? Is there any way we can make this easier or faster? faster? And just,, once a year take a look at the process. Is we need do we need to modify anything? All right, move on. Modify it. Move on. Like, can does has anything changed to allow us to be easier to build these things., and so

60:46 easier to build these things., and so I think that’s the only feedback here. Agree. This is a great use case. Use it. Patching reports are fine. Regular reports, emailing them are fine., again, what you’re doing is you’re addressing a problem and you’re solving it with some automation. The more you can automate, the happier I think your users will become. And to your point, Tommy, right, I I if if people do work in Outlook and that’s where they spend the majority of their time talking, communicating, relating to their customers, right? You

61:16 relating to their customers, right? You have to bring the work to where they are. are. Period. Period. That has to be like part of the project. Yeah. If Microsoft would give me [snorts] I don’t know why we can’t do it but if Microsoft could do a PowerBI embedded report in the email itself like literally I frame the like hey I’m logged into my email account I can open the email body and it automatically says hey there’s a pyrobi report render the report there

61:46 report there and and and log that user in because we know who you are in the email account that would be immensely like it’s almost like an iframe inside the that would be immensely valuable. That would be really cool to have that like just just working together. Now that would require I think the Outlooks team to like figure out a lot of things and like more much more deeply integrated. But how cool would it be to like open the email up and basically it’s the report that you need to have. No code. I’m not sending you images. It’s just it just takes you to what the thing is and

62:17 just takes you to what the thing is and you just work on it right there. I I I do know there’s in Outlook there’s like a button you can add an addin that lets power AI show up as like another button in Outlook which I do use I use that for me in our company but having the report interactive and working inside the email body that would be really what I want here. here. So even if I I think to I want to I want to just add on to that. I think that would be more powerful even if I had a co-pilot agent in Outlook. Yeah, Yeah, which you can actually do. to

62:48 which you can actually do. to your point, we haven’t even talked about this. If I wanted to build a data agent or a co-pilot agent and having an outlook, to me, the EML subscription as of today is more powerful because it’s proactive. The other way requires someone having to ask the question to co-pilot rather than me get getting fed the information. And again, think of a salesperson. They might not think every Monday, hey, what were all those accounts I need to contact? I get that as an email and it becomes more powerful. So yeah, I I I think yes, I agree with you

63:18 yeah, I I I think yes, I agree with you Tommy here. I fully envision that Microsoft’s going to continue have to rethinking their normal business software, right? If if I if I look at,, there’s there’s now co-pilot and there’s there’s co-pilot co-work and then there’s like cloud code. So if you look at how these things are built built in general, tools that I like to use now have much more agentic experiences on top of them. Mhm. It’s It’s not just like a hey, I threw co-pilot at your email and it’s always telling you every single time, hey, you

63:48 telling you every single time, hey, you want to rewrite this email co-pilot like I don’t really use that much. That’s not that’s not really value added for me. However, if I can build like other systems almost like take,, take your business, let me say this way. I saw a message on Twitter around this exact concept, right? Microsoft really bumbled the AI roll out for their company, right? There was two approaches, right? One was

64:19 approaches, right? One was keep your business software where where it is it is and basically stand up a new team that’s like, hey, this is a brand new team that’s totally AI and agentic focused for each tool that you have previously. How would you reimagine that in tool entirely based on based on AI? What would AI do? Like if you just say let’s just reprogram Excel and build the new version of Excel that’s only AI enabled built solutions right take your old migration software and m move it into new products that would be one way of

64:49 new products that would be one way of doing this the other way which I think is what Microsoft chose which is we already have a really solid business platform of a of a bunch of business suite and software let’s add AI to our existing things and when you do it that way it always feels like it’s an add-on it’s extra thing not as deeply integrated as you want. So until you really get to the place of like really rebuilding the software thinking okay taking a few steps back yeah to your point to what we’re talking about in this conversation let’s step back and solve the problem let’s step back to the high level and say how would

65:20 back to the high level and say how would we rethink this so it just becomes fully agentic right do we even need a list of emails anymore do we even need an inbox anymore would how would we rethink that experience in lie of the agent receives all the emails reads them all and says okay instead of you getting a a list of emails by date in your inbox. Maybe the AI just says, “Here’s a bunch of emails I read for you. Here’s some responses I’ve already curated for you. Now, I’ve already reordered them in the order you should work on them this morning. Here’s things that we think are most urgent all

65:51 things that we think are most urgent all the way down to least urgent and you could just work your way through and if you you approve and move them forward.” So, I think there’s this idea of like how would you want to integrate AI? AI? I’m me as one who builds software for companies companies thinks more about I should really probably burn down the software and start over and say okay let’s go back to core principles of what is the problem I’m trying to solve right right how do I think of that problem and solution with it in lie of AI and does the product is there a new product that

66:21 the product is there a new product that I’m coming out with that it solves the same problem but does things in a different way and I think that’s the mentality I’m I’m leaning towards because AI is changing so much of how we build. 100%. 100%. I hope we should had a game here where every time I said the word AI, someone takes a sip of soda or something like that, people would be like hopped up on coffee or whatnot this morning. morning. Yeah, I I cannot play that game for the [laughter] sake of my heart. So, all right, that being said, great question today. Subscriptions is a silent hero., unpacking what that

66:54 silent hero., unpacking what that looks like. Hope you found this session or this conversation helpful and useful to you. That being said, Tommy, where else can you find the podcast? You can find us on Apple, Spotify, or wherever your podcast. Make sure to subscribe and leave a rating. It helps us out a ton. Do you have a question, idea, or topic that you want us to talk about on a future episode? Well, head over to powerbi. tips/mpodcast. Leave your name and a great question. And finally, join us live every Tuesday and Thursday, a. m. Central, and join the conversation on all of PowerB. tips social media channels.

67:27 PowerB. tips social media channels. Thank you all so much. We appreciate your time and listening to the episode. With that being said, we hope you have a great week and we’ll see you next time. [music] Tommy and Mike lighting up the sky. Dance to the day to laugh in the mix. Fabric and I get your feels. Explicit measures. Drop the beat now. [music] Feel the crowd. Explicit measures.

67:57 Explicit [music] measures. Drop it loud.

Thank You

Want to catch us live? Join every Tuesday and Thursday at 7:30 AM Central on YouTube and LinkedIn.

Got a question? Head to powerbi.tips/empodcast and submit your topic ideas.

Listen on Spotify, Apple Podcasts, or wherever you get your podcasts.

Previous

DAX Complexity & Power Query ETL Tips - Ep. 515

More Posts

May 8, 2026

Stop Using Bookmarks - Ep.526 - Power BI tips

In Episode 526 of Explicit Measures, Mike Carlo and Tommy Puglia unpack the latest Power BI and Microsoft Fabric topics from the show. You’ll get a quick read on the episode’s biggest ideas, why they matter, and where to dig deeper in the full conversation.

May 6, 2026

Less Guessing? More Building! - Ep.525 - Power BI tips

In Episode 525 of Explicit Measures, Mike Carlo and Tommy Puglia unpack the latest Power BI and Microsoft Fabric topics from the show. You’ll get a quick read on the episode’s biggest ideas, why they matter, and where to dig deeper in the full conversation.

May 1, 2026

Fabric's Most Underrated Features - Ep.524 - Power BI tips

In Episode 524 of Explicit Measures, Mike Carlo and Tommy Puglia unpack the latest Power BI and Microsoft Fabric topics from the show. You’ll get a quick read on the episode’s biggest ideas, why they matter, and where to dig deeper in the full conversation.