Speaker 1: 0:07

Thank you very much. Hello, welcome to Microsoft Community Insight Podcast, where we share insights from community experts to stay up to date in Microsoft. My name is Nicholas and I'll be your host today. In this podcast, we will dive into Microsoft Fabric and data analytics using AI. But before we get started, I want you to follow us on social media so you never miss an episode. It helps us reach more amazing people like yourself. Today we have a special guest called Javier Villegas. Sorry if I pronounce it wrong. Could you please introduce yourself please?

Speaker 2: 1:17

Hi Nicolas. Thank you for the introduction. Happy to be here. Thank you for the invitation. My name is Javier Villegas. I am from Buenos Aires, argentina. I work with SQL Server and Data Platform since many years. I'm Microsoft MVP in the Data Platform category and well, today we're're gonna be talking a little bit about microsoft fabric okay, cool.

Speaker 1: 1:48

So could you explain to us, like, what do you do on daily basics, like as a data platform, whether it's at work or as mvp?

Speaker 2: 1:57

yeah, I mean, uh, my background is mainly sql server. I I started background is mainly SQL server. I started with SQL server 6.5 more than 30 years ago. So I've been through every single version migration, on-premise clusters, cloud, etc. And now switching basically not switching but combining with this new analytic tool, microsoft Fabric. So it's mainly everything related to the Microsoft Data Platform stack. I also part of a bunch of community from other data by station, but mainly focus on the Microsoft Stack.

Speaker 1: 2:51

Do you use Massacuff Fabric at work or just SQL?

Speaker 2: 2:55

mainly. No, absolutely. I use Fabric on a daily basis. As MVP, we have the possibility to have early access to Microsoft Fabrics, so we've been testing it myself. Personally, I've been testing it since a couple of years actually, and well, nowadays I'm using it for mission-critical analytic workflow.

Speaker 1: 3:30

Okay, brilliant. Before we dive into the fabric, we just want to start with the very basics. So what's Microsoft Fabric and what's the key components of it?

Speaker 2: 3:41

Okay, well, let's start from the beginning with a little bit of history. Right, as I said, a couple of years ago it was presented as a private preview for the data platform community right as a whole. Basically, because the key, let's say, factor of Microsoft Fabric is that it's SaaS, right, Software as a service, Meaning that you connect to Fabric through the portal and you don't have to leave it. You don't need any piece of software, you don't need any virtual machine, anything. You can do any analytical solution and we're going to dive into this through the conversation from end to end. We can ingest data from any source on-premise cloud, even different clouds. We can storage and process this data and we can create the dashboards and reports with Power BI. So that's mainly the most important thing. Back in 2023, it was presented as public review and in 2000, sorry, 2022, in 2023, during Ignite at the end of the year, it was presented as general availability, so everybody was able to jump into it.

Speaker 2: 5:29

Microsoft is giving away a lot of trial periods, so you know you want to get involved. If you would like to start testing and doing proof of concept of your own things, right, you can do it kind of for free, right? So that is super cool, right? One of the big difference also of this is, since it is a platform as a service, a software as a service, offering, right, and, as I said, you don't have to install anything to start with.

Speaker 2: 6:06

If you have a project or if you have something that you would like to validate, in the past, you had to buy the hardware, get the software, do this addition, spend some money, big money, actually to do the setup, basically to validate what you would like to do. Right, that's going to take time, effort and money, right, and then let's say, after a while, you were ready to deploy your project, right, and you should be back in the days quite careful to measure the hardware that you need for your environment very well. Otherwise, if you run out of resources processing power, storage, whatever you have to go to the vendor and buy.

Speaker 2: 7:01

With a SaaS solution like this, which you can scale at any time. You can put whatever you need as a concept. In five minutes, and let's say no more than a day, you may have your validation ready to go, and, as I said, for free, completely free right and when it's ready you can just move it to production and go to a paid offer.

Speaker 1: 7:31

So is there no cost involved, because you mentioned free?

Speaker 2: 7:36

Mm-hmm, you have 60 days trial. There are some minor things that you won't be, able to use, for example, copilot. But, as I defined it at the beginning, you will be able to ingest data, process and store data and create your Power BI report. With that you can do the validation of what you need to do. Let's say that you have to do an analytical project, for example, a normal data warehouse, right that way, you get your data from multiple sources on premise or sources in the cloud.

Speaker 2: 8:16

You don't need any connector right From what is called. You know the pipelines on the ingestion part for fabric, you connect and you get it right away For the storage part, you know, and that part of the ingestion let me go back a second is what it used to be. Let's say, let me go back a second is what it used to be. Let's say, this component in Azure that we used to have, which is the pipeline, right, data factory, right. So Fabric actually is something that is new as a concept, but it's not new in terms of the component. What do I say this? Because before Fabric right, microsoft having the data platform in Azure, you had Data Factory to create your pipelines and get the data from multiple sources. You had Azure Synapse Analytics where you can, you know, store the data using Synapse and process using T-SQL or even using Spark or Scala, and then you have Power BI to connect to the Synapse thing and present your report.

Speaker 2: 9:44

That was available, as I said, but if you think about it, it was a little bit complex to set up all these in place. Why I'm saying so? Because you know that in Azure you have to do all the connectivity, the wiring, the security. So you have three different services Data Factory, synapse and Power BI. You have to involve your system team to generate the networking, do all the links, et cetera, et cetera. Right, then another thing is that for the billing, you have to pay for the usage of the Data Factory pipeline, you have to pay for what you spin off in Synapse, even if it was a serverless or a dedicated pool, and then Power BI. So the bidding was really really complicated. Besides that, synapse used to have well, actually still have, because in fact this is still alive it has its own proprietary storage format from Microsoft. It was from Microsoft.

Speaker 2: 10:51

So now in public you can do exactly the same right All in one box. You don't need to involve any connectivity because it's all the same. The ingestion, the storage and processing and the end result with Power BI and the billing is way more simple because you have capacity. You have multiple capacities that you will choose, so you pay for that capacity. So at the end of the month you know what you're gonna pay. Yeah, it sounds like it's.

Speaker 1: 11:28

It sounds like it's all in one solution, that you just use that integrated data factory synapse and so you just go in one portal of like Max and Fabric.

Speaker 2: 11:39

Way more simple than everything. Another interesting thing portal of like Max and Fabric Way more simple than before. All right In all senses. Another interesting thing of Microsoft Fabric is the way to storage the data. I just mentioned that in Synapse, Microsoft has its own proprietary storage format for the data. Now in Fabric, Microsoft has switched to an open standard of Delta parquet, right, so doing so, it's way more easy to interact with other solutions, right, Because it's using an open standard.

Speaker 1: 12:18

Let me tell you something more.

Speaker 2: 12:21

Within Fabric, there is this component which is called OneLay, right, which is the storage part that you can use to have your data in right and share the data with different projects within the Fabric solution, right? This is just to do a simple comparison. This is like in Microsoft 365, in Microsoft Office, right? You have multiple components, like Word, excel, powerpoint, and you have one drive to store all the components and share it within. This is this similar concept, right, but for within Fabi One leg. You get the data, you store it and you can share the data with multiple projects without having to do any data duplication. That is super important because if you go back in time to where we used to have the on-premise solution, you know the on-premise solution. Let's say that within your company, you have your data set and you need to share it with a different area on your company, right? You have to make a copy, and you know what happens when you do, when you have two copies of the link and you cannot trust any of them, right?

Speaker 1: 13:35

So it's quite important to have a single one. Just in case no one's aware of it, what's OneLake? Onelake In this case, no one's aware of it.

Speaker 2: 13:43

What's One Lake? One Lake is this storage component within a fabric, in where you store the data that you get from the different sources. Right, using Delta Parquet as a storage format, right? So let's say, we will see in a minute when I present it. Right, yeah, in the storage part, right, you will be able to create multiple data storage, for example, a data warehouse, which is mainly to have relational data tables, foreign key relationship, et cetera, and you can use mainly Transact SQL to work with this. You can also create a data lake, which is basically to store any kind of data relational, semi-relational and with no schema at all.

Speaker 2: 14:39

You can put files, you can put multiple. You know JSON files, for example. You can also create tables, if you want right, and everything is at the top of one layer, right. You can also create another kind of databases, like, for example, kusto for real-time analytics. You can also store your tabular model, right, like in analysis services, and you know the internal format to store the data. Here is now delta part k. All, right, one leg also has this ability to uh, I mean, it's not only one. Like this fabric as a whole, there is this component which is called shortcuts, what you can do with shortcuts. Mainly, let's say that your company has certain data in a different cloud provider. Let's call it AWS or Google, right? Or even your own RAM data center, right? So instead of copying the data from the external cloud into OneLake, you create a shortcut and you reference that table or that data from outside without having to copy, right?

Speaker 2: 15:55

So within your project right. Let's say that you write a notebook with spot and you need to combine data from multiple sources something that is in your local length of your warehouse and through a shortcut to, for example, data in AWS. You can do it without data duplication and super fast. So this is a nice component within the cloud.

Speaker 1: 16:23

Yeah, that's quite sweet. So, regarding to like the HFAP data like Microsoft Fabric, is there any use cases? That is very good at like that. You use that work.

Speaker 2: 16:36

Yeah, absolutely you can do, as I mentioned, a traditional data warehouse right. You can also use it for data engineering right To you know, get data. Let's say you have events coming from sensors through Azure events app or real time analytics right.

Speaker 2: 17:08

You have a component which is the hub that will get those inputs right in real time and you can do your analytic at the top of it, right. For example, let's say that you have a solution that is constantly sending information about temperatures right, so you're gonna get this online right. And then, at the top of it, you can put the power bi report, which now is part of fabric right and you can do your dashboard with the real-time data right.

Speaker 1: 17:41

Does it work with App Insight as well?

Speaker 2: 17:43

Actual App Insight yeah you can connect to App Insight through Gusto, for example, and get this real quick. So I mean there are a lot of possibilities through Fabric. Recently, microsoft has announced actually not so long ago, during the last Microsoft Ignite right, they also have announced the possibility of creating an actual they call it SQL database on Fabric right, so besides creating a lake house or a data warehouse, right, you can create a SQL database on fabric right, so, besides creating a lake house or a data warehouse, right, you can create a SQL database and you can use your T-SQL knowledge to create store procedure functions, trigger right and store your relational operational data within FAB, which is super interesting. Right, that data is constantly replicating, or mirroring as we call it, into one leg. Right, and you can use this data to do analytics on your operational data without impacting the performance on your operational data. Let's say that you have an application for your clients your sales right and you have the stocks and the sales. Right, and you have your application working, creating, doing insert, update, delete on the SQL tables In near real time. Those insert, updates, delete will be synchronized into one lake and, again, you can create a Power BI dashboard at the top of it to have your analytics based on real time data.

Speaker 2: 19:28

Back in the base. We used to have what is called ETL right Extraction Transformation Load, so we connect to the operational data source once an hour, once every six hour, once a day, depending on our logic. We duplicate the data into it right, and then we build the analytic at the top of it, right? So it was not real time. We had to wait every time that we run this EDL. Now you have the possibility to do it also with real time data from operational data sources Now with SQL Server.

Speaker 2: 20:07

In the future, this might be other engines, so it's another flavor of SQL, now in power. You know that SQL is actually the part I love the most. You have the SQL Server, you have Azure SQL with Azure SQL Database and managed instance, and now you have this offering. So, as I said at the beginning, this is an end-to-end solution. It sheds the data from any source as you saw all the many sources that are available, right, store it in one lake through the lake of data warehouse or even a database, and then create your Power BI report to connect to it and show the dashboard, right? So it's all you need, okay, and, as we saw, in a minute without leaving the report.

Speaker 1: 20:59

Yeah, how can people get access to this fabric? Is it just through the website?

Speaker 2: 21:07

first of all you have to be through the uh, you have to have a, a, an entry and a group right and then, or you can assign the, the permission at the capacity level and at the workspace level. Let's say, I go back to my workspace, here, workspace01, I go here to manage access right and here I choose who can connect to the warehouse or the components underneath right.

Speaker 1: 21:45

So you can define a group here or individual accounts. That's brilliant. So how can someone get started with Fabric, Is it? What's the best places for people to learn about it if they want to learn more?

Speaker 2: 22:00

Yeah, microsoft is pushing hard on Microsoft Fabric. So the starting point is akams backslash Fabric. That's the starting point in where you can go to the option to create your trial in where you can be redirected to the Microsoft Learn page. In the Microsoft Learn page there is a nice concept called Learn Together Microsoft Fabric Edition, where that was delivered back in January, february 2024. There were like nine episodes in where it was covering everything in Fabric the basis ingestion, lake house data, warehouse, notebooks, real-time analytics, etc. Etc. Nine different episodes in English and in Spanish. It is available in YouTube, right? So you have all the training you need.

Speaker 2: 23:04

Recently we did also one quite similar for SQL databases in Fabric. Right, there are a bunch of webinars, there are books. There are a bunch of webinars, there are books. So all the resources that you need are akamsbacklashfabric. Microsoft also has created a certification for Fabric the DP600. And where several sessions on this Microsoft Learn Together initiative. We're also focusing on the exam, the DP600 exam. So there is a certification which is shared and available the DP600. Recently, microsoft is promoting a new one, which is still not fully available, which is the DP700, which is more targeting to data engineering within Fabric. So a ton of resources, free right. So you can even get several initiatives to complete training, and when you complete the training, you even gonna get discounts for the certification exam. So super cool, uh. So, if you are into the data analytics, if you would like to, you know, validate your needs, your project, in a matter of minutes you have everything you need in here.

Speaker 1: 24:35

So let's talk a bit more about the community, the data community. So I know, aside from like certifications, you can just network and be involved with the community. Can you explain more about the different community that they have in data?

Speaker 2: 24:53

Yeah, absolutely. So. You know myself, I'm from Argentina, south America. My main language is Spanish. Right, we have our own user group, sql Argentina, and where we cover the Microsoft Data Platform stack, sql, azure, sql and now Fabric as well.

Speaker 2: 25:15

There are multiple initiatives like that all around the world. There is also something quite interesting, which is the Microsoft Fabric Community Conference. Last year, microsoft did the first edition in March in Las Vegas and then at the end of the year, the same one in Europe. So now they are also promoting the Microsoft Fabric Community event for early April 2025. That's going to be the second edition in where you can learn from experts from Microsoft. You can learn from MVPs, right, and this is a community event. It's not free, definitely right, but it's the way to go. If you want to engage with the community and you want to learn more about public, right, you can do workshop on end-to-end presentation. If you are focusing in in machine learning, there's going to be workshop targeting machine learning. Same for databases and so on and so forth, right so it's the main place to to go.

Speaker 1: 26:32

Is that the one that calls fabricon or something? Is that the one that calls Fabricon?

Speaker 2: 26:36

or something Fabcom.

Speaker 1: 26:39

Fabcom. Okay, did, you went there no no, no, no.

Speaker 2: 26:43

I had some other work related activities and actually, if you remember, last year was quite close to the MVP Summit, so I went to the MVP Summit instead. Yeah, because I think it's very too close in. March, sometimes in April yeah, after the MVP summit.

Speaker 1: 27:05

So that's good. So you get to go to Vegas if it's April anyway. So because you live nearby, you live in the state yeah, I mean, unfortunately I have to.

Speaker 2: 27:17

You know, just choose one. And and you know, to me I prepared the mvp summit. I definitely have in my, my to-do list. Uh, maybe for the one in europe at the end of the year, I don't know. But but in the us, the first, first one I will be going. I have part of my team that went there. Some of them went to the first one in Las Vegas, some others to the one in Estacón for the Europe edition and they were super happy. They learned a lot. We use it on a daily basis, as I said, for production, analytic projects, right, so mission critical and well, we definitely rely a lot on all our analytics in Microsoft Parc.

Speaker 1: 28:07

Yeah, so aside from like communities and your work, what do you do? Do you have any hobbies in your spare time or you just do community?

Speaker 2: 28:18

Well, one of my hobbies is definitely community activities, right.

Speaker 1: 28:24

Yeah, same.

Speaker 2: 28:25

I like to go to different events in person, mainly to network with my friends and colleagues. But besides the technology part, I'm a dad. I have two sons, which I love a lot, and I do play tennis right. That is the way I keep my, my physique more or less into shape, uh, but mainly my, my head okay, that's brilliant.

Speaker 1: 28:58

Uh, so, before we close off this episode, I just wanted to find out are you we just made, are you going to any other events nearby in the future? Or remember you talk about you're going to your vp summit. Is that the only one that you're going to?

Speaker 2: 29:18

No, I normally go to the PaaS Summit in Seattle. This is a community event from the PaaS community. Paas is Professional Association of SQL Server. It's not a Microsoft event but since it is in Seattle, microsoft engineers go to it and they normally announce a bunch of things related to SQL and now related to all the Azure and Fabric stack. So those are my main events MVP Summit and the PaaS Summit and hopefully in the future also the FAFCon. Part of my team in Europe will also be going to SQL Beats. That this year is in June. I won't be going there. So yeah, I mean I'd love to be engaged with the community network learn. You know I love to be engaged with the community network learn. You know. The main phrase of the past community that I'm part of is connect, learn and share. Right? So definitely, this is something I'm trying to do all the time.

Speaker 1: 30:31

Yeah, that's quite. It looks like an interesting and fun community. The past one, yeah. So thanks for joining this episode, javier. Yeah, it looked like an interesting and fun community the past one, absolutely yeah. So thanks for joining this episode, javier. It's a pleasure having you. So it's in a few weeks it's going to be on social media. Stay tuned for the next episode. Thank you, bye, thank you, bye, bye.