 This is basically a business intelligence and data analysis tool based in the cloud. What I am going to talk about is how do you connect and work with business data when the business data is on-premise. When I talk about on-premise, I don't mean that your data is on a server sitting locally in your office. When I say on-premise, it means that some kind of business data which is maybe hosted in the cloud but in either a private cloud or in a data center maybe physically on-premise which is different from the cloud service itself which is Power BI. So how do you actually organize your data and how do you actually get that data into the cloud is what I am going to talk about in today's session. Just a quick introduction about myself, my name is Vinod Nih. I am a senior technology consultant in a company called Enterprise Infotech. I also happen to have a title called Microsoft Regional Director. This is a honorary title. I am not employed by Microsoft which means I don't unfortunately get a salary from Microsoft which is very sad. It's a title that Microsoft gives to people in the field whom they consider as experts. So we are about 120 in the world. So we work very closely with the Microsoft teams back in Vietnam and give them feedback and believe me we are very, very critical of Microsoft directly on their face. So when they make a mistake, we are the ones who go and tell them that boss you've really screwed up. So that's a quick introduction about myself and some other stuff on the slide where I am going to skip through that. So quick agenda what I am going to talk about. So first I am not sure whether it's very clear but don't worry about that. So I will give you an overview of what VI actually is and what are the different solutions that are there and how do you actually work with business intelligence. And then we will take a look at Power BI and how it actually helps you solve this in the cloud and get that data from your business data, your business processes, your DevOps processes and push them into the cloud to do analysis on them and not the advantage of doing all of that. So that's what I am going to talk about and I will try and show you a few live demos if time permits and if the internet bots are on our side. I will try and show a few demos live as well and live demos are always dangerous. They sometimes won't work. So when I will attempt it, I have taken a video of it working. So in case it doesn't work, I will show you the video of it working. So a quick run through of what is VI, how many of you work with VI? None of you, right? So VI, if you look at business intelligence and you read the books that you read about it, there is a very complex definition of VI. You can see by, and there is just one sentence that I have taken out of the official definition of business intelligence. It's called a broad category of application programs and technologies for gathering, storing, blah, blah, blah, blah, blah, blah, blah. It doesn't make sense. To put it a little bit more simply, I will put it a little bit more in normal English. It is the application of knowledge derived from analyzing an organization's data to set a more positive outcome. Where it's actually a little simpler, basically knowledge by analyzing data to get some good positive result to put it even more simply. It is basically picking up data, raw data, converting it to information and then from that information getting some knowledge out of it. That's what business intelligence is all about. So forget all the stuff that you will read about. If you ever get into VI, you will read all this stuff. But the essence, and VI is an old technology, it's been around for 30, 40 years. The tools are new, the way that you evolve and use it is new but the concepts behind it is not very, it's been around for a long time. But in the end, the essence is basically data information to knowledge. That's what you need to understand what business intelligence is. Obviously with the tools that are there and available today in the cloud and stuff like that, there are a whole lot of new ways that you can get all of this information. So what is VI used for? It's used for understanding the health of the organization. This could mean anything. This could mean how are sales doing. How are sales doing in a particular region? How are your servers health doing? It could be any kind of data. It does not matter what that data is. The VI is used to actually analyze, understand and then even predict in some cases what is going to happen. I was going to ask you on a shared view of business drivers. Now one of the problems that happens when you have a large organization is that you might not have the same view of business drivers. For example, if a sales region in India sells items in rupees and a sales region in the US sells it in dollars, a $10 sale is not equivalent to a $10 sale. You need to have a common view. That means you need to convert it into a common set, a common unit, a common thing. You need to collaborate and view that shared data driver and then of course reduce the time of decision. You need to have information extremely fast and I'll show you how that happens if the demo works fine. And obviously the goal is to impact the bottom line. That means it doesn't matter whether your startup has just shipped its first item today. It doesn't matter. Or if you are a 30 year old company multinational with tens of thousands of employees, as long as you are able to analyze the data and reach a decision quickly, your product might be an app and how is that app doing? How your product might be t-shirts and shoes and how is that selling? Or your product might be services and how are your servers doing? Now I don't mean real-time server, real-time information of whether a server is down or up. That you get from AWS and Azure and all of them, you get that. But how have they been performing over time? What are the issues that have come up? What issues affect what? The different types of insights that can come up are quite important to understand. So now how do you do all of this? What are the BI solutions that are available there? So the problem in today's world is that organizations have data stored in many different areas. There are silos of data. Data could be on server logs. Data could be in databases. Data could be in CRNs, ERP systems. Lots of different things that are available. And they may not be optimized for analytical queries. Now what does that mean? How do you analyze this data? If you look at a server log, I mean if you look at raw server logs, you can really never understand what is there. If you look at raw data in a database of sales, transactional data, you can never understand what is actually going on. What is actually happening? You are going to be at individual information at one time. How do you manage historical context? What has things been in the past? How is it happening now? What is the trend? So that is not available as you look at data as you go along. The other problem is that most employees don't have the skills to analyze data. Most people don't know how to do it. In fact, BI is a very technical concept. You need to have understanding of data models and a lot of other things. So that is not something that a business manager or decision maker will actually have to understand. So that's where BI actually comes in. If you look at data silos, they could be as I said, it could be multiple data silos. HR, finance, the IRM, these are all silos of your organization. If you look at silos that are coming in from outside your organization, there could be your vendor's data. How do you correlate your data with the vendor's data? Or how do you correlate your data with external data? For example, external data could be things like weather information. Is your sale actually impacted if the weather is bad? How do you figure that out? That is where this sort of stuff comes in. Where you can actually go into business intelligence processing and then come out with analytics and visualizations which tell you this information that are not insights that can come in. Now, business intelligence, there are tons of solutions that are out there. I am going to talk about one which is called PowerPlayer, which is basically a suite of business analytical tools from Microsoft. Now, just because this comes from Microsoft does not mean that it only works with, say, SQL server. It is completely platform agnostic. It does not matter. Your data could be anything, MySQL, Mongo, Cosmos, whatever, it does not matter. Whether it is stored locally on your laptop, it does not matter. It could actually be physically on your laptop, not even on server for all that matters and you are sitting inside your office and you could still get that data analyzed. PowerPlayer itself is available, so it is a cloud-based service. It is available both stat alone and you can get it as part of Office 365, so it depends how you want to get it. It connects, as I said, to hundreds of different types of data sources and they are adding more data sources on a day-to-day basis almost. The data thread, which is the more complicated part which usually requires a consultant to do it or you need somebody with business intelligence knowledge to be able to do that, that is done via a very simple user experience. I will show that to you in a moment, very, very simple to do. If you have worked with Excel, you basically know how to use it, more or less. And you can do extremely detailed queries via these prepared things or completely ad hoc. If you do not know how to do it, you can do a very ad hoc query and still get data analysis done on that. The cloud-based service is actually, because it is backed by Microsoft's cloud it has AI and NLP went in, which means that a non-technical user like a manager can actually go and type in a query. Tell me the sales, you know, give me an analysis of sales of France over the last three years. And you will get a proper visualization from the Power BI service which will actually show that information very clearly out. Okay, this is the entire sale history of France. What are the regions that are there? So very, very customizable and very nice interactive and fluid visualizations that are there. Now, how is Power BI available? There are two versions. There is a desktop version and there is a service. The service is obviously the most powerful thing. The desktop is actually just creating the user experience and service itself is available in three different subscriptions for personal, pro and premium. As you quickly run through this, the desktop is obviously on the desktop. It's free. You can just simply go and download it from Power BI.com. It's also available on the Windows. So just simply go download, install and you can start using it. Power BI Personal is again free. It's a cloud-based service. It's absolutely free. You can actually go and start using it. The only restriction here in Power BI Personal is that any analysis that you do cannot be shared with others in your organization. That's the only restriction. Otherwise, everything else is there. Power BI Pro, which is the full edition of Power BI, again available standalone or available as Modoc Office 365, allows, so the difference between the personal pro, both are cloud-based, but pro allows you to share any analysis that you've done within your organization, just with a couple of clips. And this costs about 600 rupees per user per month, which is actually very, very cheap. So even a startup can afford this. You need one Power BI Pro license and 600 rupees per month. That's what we need. Power BI Premium is obviously the premium experience. This is expensive. It has a capacity plan, cloud-based service, and available via a license partner only. This starts from 6 lakh rupees per month. That's the first level. So this is equivalent to, if you've heard of other platforms like Tabu. Tabu is the direct competitor to Power BI Premium because they start at around the same price. And that is, of course, your break-even is around when you have 500 users of Power BI. That's when your break-even starts off. So if you are fired on that, then, of course, Power BI Premium is a very good thing. But before that, it's kind of very good. So how do you analyze data? Very, very simple. You basically connect or point your Power BI service to your data source. The data source, as I said, can be anywhere. It could be in the cloud. It could be inside a compromised network. It could be in your data center. It could be on your laptop for all it matters. It doesn't matter. It will connect to that. You create a data model based on the data source, the data that you're trying to analyze. You create certain reports. You apply some visualizations, and there are some things. And I'll try and show this to you. You highlight, so if you have a report with some 20 different insights that you are working with, and you want to highlight some of them, you can highlight them into a dashboard, which maybe managers or somebody can quickly take a look at those dashboards to see what's happening. And now go share the report on the dashboard or the content itself to others in your organization. I'm just going to try and show you a quick demo of this. So what I have here, and I'm just going to quickly go to the cloud-based service. So I'm logging in. So this is already connected to my, so I'm connected to the Power BI service. This is one report that I have actually created. It's not probably going to be very clear because of the resolution that is showing up. But this is some sample sales data that is there. The colors actually have much more vibrant results showing up here. So this is basically sales over time, right? So this is sales over time and by the region. This one is sales over the category. So there are different categories. This is sales over currency. This is sales over region. So you can see North America. Right now, within that you have United States and Canada. You have Europe. You have United Kingdom, Germany and France. And this is again sales over time. Right now, this looks like a simple enough report which you can quickly create in Excel. But the power flag comes up is that one it is shareable. The second is that it's cross-highlightable. For example, if I click on United States dollar here. What happens is that I don't really notice but the values on the graphs actually change based on what are selections. So you can see that these are what highlighted. So Canada has fallen. Canada was full. But the highlighted part is only this much. And these graphs have actually all gotten changed based on what you've gotten selected. And it's cross-highlightable in any direction. So I can actually go and say, okay, I want to see only for United Kingdom. I only want to see for Germany. Or I can say, okay, control click both. And I can see what is the combined cumulative effect of Germany. So that's one quick way of creating reports. Reports can have multiple pages. There are more pages here. So I can just quickly show that to you. More pages which I've just created as a sample. Now these are reports. You also have a dashboard. So a dashboard is basically where I've taken some parts of my report and said, I want to show this as a quick dashboard so that somebody can get a quick look at what this is. And here if you notice there's something to ask a question about your data. This is the energy part. I'm not going to demo that because I'm not set it up for that. The other thing which is very nice, and I remember that I said that for non-technical users, the ability to access this information needs to be very intuitive. So at least for now it's on Windows 10, but you also have Power BI apps for all platforms. iOS, Android, Windows, Mac, everything. So this I'm of course showing on a Windows 10 machine. And there's one intuitive thing here is that if you're using the speech assistant, you can actually go and talk to your speech assistant here on Windows 10 and ask me to show you this report. So I don't need this window open. I can actually go and type in it out, but I can actually speak it out as well. So I just go and say Power BI Sales Report. And you'll notice that the sales report comes up here. I click on it and it doesn't open a website for me. It actually opens that dashboard right here with all the possibilities that I can actually do. So I don't even need to go to a website to actually go and do things like this. And you can see the numbers are changing. So it becomes a live dashboard right there in the Windows user interface itself. You can do this on Android and iOS also as long as you install the Cortana app on that. And of course you have a pre-built app, a specific app for this as well. So that's a quick run-through of Power BI because that's not really what I'm talking about here. So now the question comes there, how do I get access to this data? My data could be anywhere right? Let's say that data is sitting on a server on my premises. It could be in a data center, but for all practical purposes that's on premises. So Power BI gives something called the on-premise data gateway. This is a kind of service that gets an agent which gets installed within an environment. I won't call it on the same server because it can actually handle multiple servers. It gets installed somewhere in your environment where the machine on which you install the gateway has access to all your data sources. Once it's installed within the network parameter, the data gateway has access to the data sources and that will then allow Power BI service to actually access your data. Now the data transfer itself is completely compressed and it is encrypted so the data on in transit is secure. So and the data on rest is also secure on the Power BI service. So that you don't have to worry about. So you can actually go ahead and query any kind of source data very very easily. Now the only requirements for this gateway is that the machine has to be on a Power BI server running 64 bit windows. It could actually be even on a Windows 10 machine or even a Windows 7 or 8 machine as long as it's a 64 bit operating system and it's always on for all practical purposes. You can install multiple gateways within the same network if you have that many data sources that you want to access you can actually do it but it should be as close as possible to the data source. So if you have a database server, let's say you have a MySQL database sitting somewhere in your device in your data center. Install the gateway on a server either on the same server or on a server which is very close in terms of network. It's as close as possible to the data source itself. By default it uses outbound TCP 443 which is a secure TCP by default but in case TCP is disabled, any kind of TCP transmission is disabled you can configure to use HTTPS also. However there will be a small performance penalty in the case that you do that. Now the architecture is this, this is how it works. So you have a Power BI service sitting somewhere here you have something called a gateway cloud service which is hosted again in the cloud you don't have to worry about any of these other things. I'll come to this in a moment. You have a permitted network here and you install the gateway service on a machine somewhere in this network which has access to all your other data sources. So it could be databases, it could be even files. It could be Excel files, it could be access files, it could be anything. Basically anything which could be used as a data source it could be your analysis cubes if you have, if you already have created analysis cubes you can actually go ahead and do that. As long as the gateway has access to one of them your Power BI service will also get access. Now one of the nice things about this gateway service is that this gateway service is actually available not just to Power BI but for a bunch of other services as well. So there are things like Power apps and there's analysis services there's logic apps. So if you're writing AI based systems which need to access your data which your organization's data and take decisions and respond to users based on that data you can actually have that work with the same data gateway. You don't have separate data gateways for each of them. It just works very, very well with all of them. Flow is if anybody use IFTTT, right? If you use IFTTT if this then that which is a very nice way of you know kind of getting cloudware services from each other. Flow is an enterprise version of the same thing where you can have if something happens then do that kind of thing based on the data in your organization, right? So you can actually write, I mean not write code it's graphical for all practical purposes. You can actually go in and do that and it works quite nicely. So how does the data gateway work? It allows access to your organization data in two different modes. One is direct query where the moment I showed you a report every time a report is generated the report actually goes, talks to the gateway and then goes and talks to the actual data source and pulls the data out at that point of time, right? So it's sent every time and it's pulled, the data is pulled so it's very, it's real time at that point of time. The moment you look at it, it's real time. Obviously you should only use it when you want that level of real time data because obviously there will be a little bit of lag when that happens. Your gateway server also needs to be that much more powerful to be able to access this. The other is something called scheduled refresh where you can set up a schedule and say that okay I want my data to be refreshed on an hourly basis, dk basis, vk basis or whatever it is and the gateway will be responsible for actually pulling that data out, operating the visualizations and then going and showing all of that. So the configuration is very, very simple. There are two levels that you need to configure one is on the server, the gateway server itself and the second is configured on the Power BI cloud service so the screenshots are not clear. So what I have, I've already done this so I'm just going to quickly run into a small demo of this because I actually already did this. So what happens is when you're installing the Power BI gateway, this is where you actually go ahead and do it. You actually go in your report, go download the gateway run it on the server that you're trying to run it. It's very, very simple. Just say next, next, next. So there are two versions but obviously the recommended versions what you choose, there's also something about personal which I'm not ready to get into because that's meant for, as I said, if you want it from your laptop, if you want to send some data on Power BI you can do that. That's using a personal gateway. But in the case of an enterprise, that's basically what you need. So the first time you run it, the gateway actually goes and says that it obviously says where you want to install it and once it is installed it just makes it, and this is real time, right? I've not sped it up. This is what is happening in real time. It's actually just installing. I had done this on the server and the data that you saw is actually coming from the server sitting somewhere. And once it's done, for your cloud-based credentials for Power BI. So if you have a Power BI license, you basically give the cloud service credentials and it goes and configures the gateway for you at that point of time. So on the server side, that's basically all that you need to code. So that's the email address to use with this gateway. That's it, right? So you just basically give it. That's my email ID that I'm actually typing in. And ask me for the credentials. I enter my password. Obviously the name has to be given. What is the gateway name? Because you could have multiple gateways within your organization. So I'm just giving samples gateway or something like that and a recovery key. So in case you want to migrate your gateway from a lesser powerful machine to a more powerful machine, but you don't want to reconfigure all your gateways, you give the recovery key. So you install in your server, install this and say, okay, I want to recover from your existing gateway. Just give this password or the recovery key and it reconfigures it based on the existing one and decommission to the old one. So your old server gets decommission and that's it. That's done. That's the configuration all that you need. And the configuration is done. You can see it's ready, ready. The logic app requires a little bit more configuration but for Power BI, it's kind of done. And the only configuration that you could change is this one, the HTTPS mode. So in case your system is not able to connect so each state is connected, if it's not able to connect to the Power BI server, it will say not connected. So this one, network status connected. If it's not connected, it will show up in red and say that it's not connected. In that case, just switch this HTTPS mode on and then retry it should work. That's basically it. On the server end, on the cloud end, there's a little bit more configuration that's required. So what you do is you go to the cloud, you go to the report. So this is the report. So you go to the data set. Basically the data set is the place, the information that you're pulling out. And you need to say that, okay, I want to pull this out using this gateway. So what you do is first of all, go to the gateway. So I'll just show you. So you go to the data sets. You quickly go to the, I can show this live as well. So here it says you don't have a gateway installed because I've already configured it on the cloud. So I can go and say manage gateways. So there's a link which says manage gateways and you'll see that the gateway actually shows up here. So this is the gateway that I just finished installing. It's Apple's gateway. And now I just go and add a data source. So my gateway is installed. Now I need to tell the gateway, hey, where do I need to connect to? So I go and say, okay, in my case it was a SQL server machine. So I give it the name of the server. I give it the credentials to that server. So I'm going to jump through this. So I'm giving some credentials. And that's basically it. I can see, so now I go to the data set and say, okay, for this particular report, I want to use this gateway, right? So now you can see that this says online. This is the gateway that is there. And you will go and just turn this on. So you use a gateway, apply. And it's already picked up the data source credentials automatically. Say that, okay, fine. I understand that this data is coming from this gateway and this is the credentials that is there. That's basically it. And now I can say I want to schedule. I want a refresh, which is scheduled. So I can actually go and pick up the data on a daily basis at some particular piece, particular time. I can do it multiple times a day. So the schedule refresh is up to one hour. Right? So I can have refreshes up to one hour. Direct queries are instantaneous. I mean, it's real time. Right? So that's basically it. It's as simple as going ahead and setting this up and setting up your gateway or to connect to a data source. So that's quick run through of all we have. So if you have any questions, I'll take that up. And that's my comment. Thank you.