 Hello and welcome. My name is Shannon Kemp and I'm the Executive Editor of DataVersity. We'd like to thank you for joining this DataVersity webinar, Improve Your Sales Analytics with Alteryx at Microsoft Power BI, sponsored today by Alteryx. Just a couple of points to get us started. Due to the large number of people that attend these sessions, you will be muted during the webinar. For questions, we'll be collecting them via the Q&A in the bottom right hand corner of your screen. And if you'd like to tweet, we encourage you to share highlights or questions by Twitter using hashtag DataVersity. If you'd like to chat with us or with each other, we certainly encourage you to do so. Just click the chat icon in the upper right for that feature. And as always, we will send a follow-up email within two business days containing links to the slides, the recording of this session, and additional information requested throughout the webinar. Now, let me introduce our speakers for today, Miguel Martinez, Dan Gananchal, and Ali Said. Dan Gananchal is an Alliance Marketing Manager at Alteryx. Dan currently focuses on partner marketing efforts with Microsoft. He has held several roles in his career where he has produced a strong record in driving both customer and revenue growth. Miguel is a Senior Product Marketing Manager for Microsoft Power BI. He oversees the digital strategy for the cloud business intelligence solution. His background includes formal training and practical experience in marketing and engineering. Ali is a Solutions Engineer at Alteryx. Ali has spent three years with Alteryx providing customers with solutions to their data and analytic needs, ranging from data preparation to predictive modeling. And with that, let me turn it over to Dan to get us started. Hello and welcome. Thanks, Shannon, for that introduction. Definitely appreciate it. Just so everyone knows, we're very excited for today's webinar, and we're also excited to be presenting along with Microsoft. And I know we have many people here today that are data and sales professionals, so welcome. Your work with sales data is crucial to organization success, and because of this fact, it is very important for you to identify ways in which you can optimize your work when it comes to sales analytics. And so our goal in today's webinar is to show you how you can reduce the time and complexity spent when working with sales data. Basically, we want you to see and learn how you can spend less time getting your data put together, prepared for analysis, and then visualized for insights by reducing the overall amount of time it takes you to go through the data process. And so before we jump into talking about what that looks like and what that involves, let me go ahead and briefly introduce to you the technologies that we will be using to further explain and show you how we go through the data process. And the two technologies that we'll be using is obviously Altrix as well as Microsoft Power BI. And so in terms of Altrix, just a quick introduction of that so everyone's aware. Altrix is the leading platform for self-service analytics. Basically, Altrix Analytics provides analysts with a unique ability to easily prep, blend, and analyze data using a repeatable workflow and then deploy and share those analytics at scale for deeper insights and hours instead of weeks. The Altrix Analytics platform, it can connect to and cleanse data from data warehouses, cloud applications, spreadsheets, and other sources, easily join that data together, blend it together, then perform analytics such as predictive, statistical, spatial using an easy and intuitive user interface. No coding necessary. You'll see this when we get into the demo and draft and drop workflow and how it's really easy to build it all together. So again, Ali will go into more detail about that when we show you the demo of how we use Altrix to improve self-analytics. As far as Power BI, Miguel, would you kindly give us an introduction on Power BI? Of course, Dan. Thank you. And hi, everyone. For those of you who don't know what Power BI is up to now, it's basically Microsoft offering for business intelligence, data analytics, and data visualization. As you can see here, we have a very simple diagram that shows all the data you can connect to, including, of course, Altrix directly into Power BI. And just to give you an idea of the short period of time that Power BI has been out in the market and the success that it had, it has over five million subscribers in 210 countries and more than 200,000 organizations. And actually, this data is from March this year. As you can imagine, this keeps going and going. The key components of Power BI are, of course, it's Azure-based, so it runs in Azure. It's a browser-based. It has all mobile apps on any platform that you can think of. And also it has a companion app called Power BI Desktop. So we cover all the analytics process from tapping into data, modeling it, visualizing it, and then, of course, consuming it, which is the biggest part of Power BI. And that's what we're going to see today, taking data from Altrix using all of their goodness. Thank you, Dan. Cool. Did you want me to go through these other slides or I guess we could just pass forward through them? Yeah, let's look forward to them. So just the biggest components, you're going to see dashboards, which is the main component of Power BI. That's where you can choose what you want to highlight from the data that you have. Once you click on those tiles on the dashboard, you're going to see the update, and that's going to be much more clear in the demo. The things that we're trying to solve with Power BI, the data challenges that you see, especially in sales organizations, like it's a very slow process, time-consuming to clean up the data. That's where Altrix comes in and also share the data and keep it in one source of the truth, and that's where Power BI gets in. Tactical, meaning it's very easy to share. So you solve that problem with poor visibility of the same KPI and the requirement of IT, if you want, like on the old BI, so to call it, that's where you go and you ask for a report and you get it back or you get access to a data and you analyze it in Excel. Power BI is trying to solve for that, again, keeping that data at one single location, with everyone seeing that same source, but at the same time flexible enough that you can ask your own questions. If you can move to the next one. I think I already answered this. This is how Power BI tackles those which is accelerating very slow. Transforming being more strategic real-time, getting access to that data by the second and work together more effectively. Next? Thanks, we go. Oh, thank you, there. All right, cool. So now that you've been introduced to us and that's knowledge that you've been using today, let's go ahead and jump into the presentation. What we're going to be discussing is we're going to explore three ways in which to get settled insights quicker with more efficient and effective preparation and visualization of your data. Specifically, we will discuss how to improve sales forecasting by easily blending all your relevant sales data from multiple sources. In a tenth of the time, it would probably usually take you to do it manually, or do other more cumbersome and complex processes. Two, we'll explore how to gain deeper insights into your sales data by utilizing predictive analysis with no coding required. And three, how to visualize your sales data to uncover data. And to visualize your sales data to uncover trends and relationships that can help you with customer profiling and targeting offers. So to begin, let's talk about the overall data process first. So analysts, like yourself, depend on massive amounts of data to produce actionable insights so that your decision makers can quickly react to the ever-shifting customer demands and business needs. And to keep up with the pace of business, you rely on this data to drive deeper insights that help answer strategic questions. Now, sales data comes in many forms and from multiple sources. We all know that there's a lot of things that you can get sales data from. But that data can be messy, incomplete, and a lot of times difficult to process. And oftentimes, the data requires significant manual preparation and loading and clouncing before you can analyze the results and deliver answers to your organization in a timely manner. So it's no secret that analysts spend significant amounts of time letting and repairing data for analysis and reporting. In addition, you might have to enrich the data as well as perform advanced analytics like predictive or geospatial analysis to generate deeper insights into what your data is telling you. Then you need to analyze, share, and visualize the data to help your decision makers better understand the insights in order to react to business needs and customer demands. So typically, this entire data preparation and process could take you up to several weeks to perform, even months if you have a big, large project and you're dealing with large complex data sets. The problem is that in today's space world, spending weeks to generate insights is not ideal. And so we took a survey of organizations and as you can see here, spending weeks to generate insights cause delays and these delays are causing real business problems. One such problem is business opportunities, which account for 37% of the total business area to impact the most versus any other reason on here. So losing out on potential sales opportunities definitely has a very negative effect on your business. We also learned that organizations want insight in hours or less. Reacting rapidly is very important as decision makers strive to reduce sales opportunities. Hence you need a better way to quickly visualize your data to produce deeper insights, again, less speeding up your time to insights. So to give you an idea of how these two are being addressed, we're going to use an example, a use case from Mazda who is actually using both technologies in conjunction to address their sales analytics problems. To give you a brief background, Mazda Motor Corporation they presented at a recent inspire conference and so let's take a look at what they're doing. So Mazda has been around since 1920 employed 44,000 employees globally. The company headquarters in urban California and if you're not aware of the car they produced, they produced such automobiles as the popular RX-7 and the Miata. At Mazda their group manager of retail programs and CPO for North America faced several key business intelligence challenges while he and his team were working with sales data. So some of those challenges were analysts had to hard code hard key PDFs to excel in order to be able to analyze the data from those PDFs and this generated a lot of static reports and combined reports in Excel. Analyst would also receive reports in different formats. He mentioned that for most of the time analysts would take up to 8 hours per week on their own just to make one combined report for the execs. With all the different reports floating around there were multiple versions of the truth amongst the department and so cross-functional teams would spend hours and meetings arguing about what a sale quote unquote a sale was and how it should be classified in various reports. Furthermore there were not a lot of data people that could take a problem and think about it in a database or a data analytics type of way hence doing data analysis was not always easy. So lots of time and resources were spent on these issues. Now entered in Ultrix and Microsoft Power BI that they brought in to help them solve these issues and reduce the time spent from analysis to insight. Basically what was done is Ultrix enabled Mazda to reduce the time spent in data blending and preparation by eliminating manual processes and complexity to perform fast and dynamic matching across data sources for new insights and then three, eliminate delays caused by waiting on others with the technical R coding skills needed to perform predictive analysis on the data. Essentially bridging the gap between a non-technical analyst skills and the technical knowledge needed or required to perform predictive analytics. As far as Microsoft Power BI is concerned that empowered Mazda to gain insights very quickly into what was going on in their sales territories and their sales regions by allowing them to export their sales data to Power BI for manipulation of visualizations instantly. So let's take a look at what Mazda did. So the first thing is Mazda wanted to know how many retail vehicles were sold by market and I'm sure this is probably something common that a lot of sales organizations face. Like how do we drill down deeper to our data to understand what we're selling by region, by month, by market, whatever the parameters are. So Mazda combined that data with data on what competitors were spending on incenses and this gave them a clear view of what happened one to two months ago before this through a lot of anecdotes. For example, so and so said that this dealer had a bad month because of weather, so and so said that this dealer really sold a lot because of this reason but no one really actually knew that they could back up the real reason. So Mazda wanted to take a look and now with the technologies Mazda could look very carefully at what happened in specific markets or what actions competitors took to drive outcomes. Outcomes that provided insights why Mazda was under or overperforming in specific markets. Another thing that they wanted to do is they developed a regression model to tell them on any given day of the month how many cars they would sell. So Mazda was done on a gut feel but this new information was then used by the accounting team to help with determining cash flow forecast and Mazda's sales team also used this information to determine how many cars their region should sell on any given day. So this enabled corporate to monitor if their cars were on track or off track to sell versus their objective based on the day of the month. For example, if they knew that three, which is one of their cars, should sell about a thousand units by the 15th day of the month and after looking at data it is not selling that quickly then that obviously told them there's a problem to be something that's going on that they need to investigate and see let's further drill down into data to see what that might be. So it got them instead of trying to figure things out on a gut feel it got them more clarity into why they were under or overperforming. So by combining the power of altrix and power BI Mazda was now able to one, look at the industry from a new perspective by combining data streams to predict the future or at least have a reliable benchmark to assess performance and three, focus on quote unquote the next problem not recurring problems. Now you're all probably wondering well how did they do this? So to show you how Mazda utilized altrix and power BI let's go ahead and demo some similar examples. We can't actually show you what they did because that's confidential but we'll show you some similar examples of how you can easily blend all your relevant cell data from multiple sources, utilize predictive analysis with no coding, and then visualize your cell data to gain deeper insights. And we'll start off with Miguel. If Miguel first you could show us the visualization piece that you would have used. Of course, and thank you. So it's good to show the end product before we actually go into power cell. Let me share my screen. Let me know when we can see it. Here we go. It says that I'm sharing so please let us know if everyone can see it. Perfect. So what you see here is the main screen of power BI. Or you have your dashboard which is that over level view of your business. You have your reports which are the drill downs to all the data that you have and you have the data set. All of these components have several features that make analysis and data visualization very, very interesting. What you will get out of altrix on the demo that we're going to see after is basically a data set. So let me show you what the power BI will do. We'll do something very simple. Let me go to here. As you can see, you have the blank gambas. You can see all the data available from that data model in particular. We can do things as simple as search for sales and go with net sales, drag and drop. And then we'll see if there's a function that data set also which hopefully there will be. Very simple data visualization. Of course, the idea here is that you can represent the data any way you want it. So because this is geographical data, I can click them up and that comes from the model. Different types of maps. We have some very cool custom visualizations that you can find on this gallery. There's 10 of them. So I actually just have one just to show you what that looks like. You import the custom visualization, I click on it and you should get this very cool 3D like which again is not very useful in this particular case, but you can see what are the effects or the different types of visualizations that you can get. Once you're happy with your visualization, you will get to a place with something like this. These are the reports. Data visualization class reports. Here's where you have all the details, you have slicers, different types of visualizations. Each page in this case represents one of the different topics that we wanted to take a look at. In this example, you're seeing a pipeline report where you can see all the opportunities in each stage. What are the different opportunities by account, do you want to go and choose what you want to pin to your dashboard? The dashboard again, remember, is that over version or vision of what you want to share with your colleagues or with other people. The way you pin something to a dashboard is simply clicking on that in the icon that you see. When you click on that, that's where you choose to send something to the dashboard. We start playing around with this a little bit and see what happens. One of the coolest features of Power BI, and if you remember I mentioned before when we were describing the general view of Power BI, is that we want to make data accessible to everyone. The fact that you have an interactive visualization, it's part of the deal, but we want to bring data to everyone. Even people that don't know how to handle that don't have the Q&A. You have some recommendations here or some recommendations, so what you can ask, the idea is that you can ask in natural language things like, for example, margin by product. And then you can start forcing this, let's do it by product category, I guess. You can recommend recommendations as trim out. You can pose the type of visualization to. Let's go with margin by product category mix. Let's map and see if I can push it that way. And then you can get a visualization that is basically representing all the things that you wanted. This is full interactive. If you like what you're seeing you can always pin this back to your dashboard. The moment people click on that tile in particular, they will give you the response and the data in that way. Let's get a little bit more tricky here. We want to know how we're doing versus target. We want to know why. For example, one great opportunity would be we have $16 million in the negotiation phase which is very close to being added to actual revenue. We can actually drill down on the top and then when you actually interact with this visualization, you're going to see that we drill down to the next level of that hierarchy which comes from the model that you can build without trade. Here you can see all the sellers and you can see who actually has the more dollars in that particular stage. If I click on there, you can actually see that the account is salvage video which would have moved forward because this represents a huge chunk of the sales bill one. If I go back up, we can do things like, for example, I want to focus on the account type that is planning them. If I click on planning them, you're going to see all the data being updated. That could be a huge opportunity for doing that. We have things like slicers that you can see on the screen. Wrong time. So we have a classic variance budget like country. If we focus on Germany, which is the most negative one here, you can see that there's a huge difference, for example, in the target mix for electronic services, the actual ones that we're selling. You can start drilling down the same way we did before. You can actually find the root data here. Last but not least, going back to that first thing that I mentioned, let's say you have the data set that you imported and you want to know quick analysis of the things that are related to each other. We have a cool feature called quick insights. You go to that data set, you can see that it's working right here and basically what the Power BI is doing is let's say quick stats on what's related to what. That comes in a way of a summarized table that we can check right now. These are the same tiles that I mentioned before. You can get what's influencing from the product point of view, the average financial budget, the actual dollars financial budget, so on and so forth. If you want something back to your dashboard, and each tile actually will give you the option of doing those quick insights just particularly to the data that you're seeing on that tile. I think that's it then. Those are very fast, but I think very complete overview of Power BI. I'll be answering questions from the Q&A tab if you guys have any more specific questions around Power BI. Go to dashboard. Of course, let me do that right now. The way this works is you're going to see a menu that shows you which dashboard you want to record, and let's choose the random ones here. Using your question, I'm going to show something else. As you can imagine, Power BI takes a lot of data sources. In this case, I have an Excel report. The way you do that, let me go to the report itself. Every visualization within the report, let me go to Python reports here. When you hover over any of these particular tiles that you see, any of these particular visualizations in the report, you're going to see on top this in Python. When I click on it, you're going to get a menu that tells you where do you want to see this particular visualization. In this case, I'm going to go sales dashboard. I'm going to say pin. I get the message that pin is successful. I'm going to go to sales dashboard, which is the one shown before. You can see the tile here. The way this works is you can move it around that you have full flexibility to what you want to see. I'm going to show you where the art is from the data analytics point of view. Showing the things you want to highlight when somebody wants to build down on something, that's the action of clicking on the tile, going to that particular report and displaying the information that you want in that report. Did I answer your question, Dan? And I'll be answering the question while we keep on the presentation. We'll have some time at the end. That's the visualization part, but obviously to get to visualization, first we need to prepare, cleanse, blend the data to get it ready for us inside. I'm going to hand it over to Ali. If you could show us how that part of the data process started. Go ahead and share my screen. So, sharing my screen and what you see right now is the altrix designer that is sitting locally on my desktop. So we're going to use altrix to build a workflow to do some data preparation, some aggregation and predictive analytics to prep a data set to then send to Power BI that we can then do some visualization on. So the altrix designer is a workflow-based tool. So we'll build our workflow on this canvas using the tools that are above the canvas. So you'll see that the tools are separated into groups or categories based on functionality. So we'll see tools to connect to data sources or databases, different tools for data preparation. Whether that's cleansing data which I'm going to show you, I know how to take out white space, applying a formula so we can union or stack data sets on top of each other. We can join data sets together so analogous to doing a VLOOKUP or even do fuzzy matching. So that allows us to match strings that may not have the same expelling. Different ways to parse data so if you've got log data, unstructured data you're pulling from the web, we can always parse it using web expressions or text columns into a table. The transform category of tools where you'll see tools allow us to aggregate the data and we'll move into the more advanced analytics category of tools. If you're doing spatial processing, mapping, point to point, you have that available to do in altrux as well. And then the statistical piece. So we've built in R into the tool so you can do different kinds of predictive models, maybe such as logistic regression, linear regression force models but you don't have to do any coding, you're simply making selections via GUI. I'm going to show you this in a time series model for our workflow. So let's go ahead and get started. We're generally going to get started by connecting to the data set. So I can drag and drop a tool into the canvas. Notice that I do that, the configuration window for that tool changes so I configure this specific tool. So we're first going to connect to data, we're first going to connect to a local file. Some of the file browse. To give you an example of the files that we can pull in, you'll notice that we can pull in Microsoft Excel, Microsoft Access, even on structured text files. So I'm first going to connect to a CSV. So I made that data connection. And this CSV has data about our sales. So on each day, you know, transaction level data, it shows us a pretty transaction what was the value of each of those sales. So I just pulled in one file but also got some other capability as well. So you'll notice that there is about nine different files there. So if I wanted to actually pull in and have altrix automatically union them for me, I could do so by simply placing the wild card here. So instead of just taking that sales in a square one file, take every file in that folder that actually starts with sales and I could just place a star right there which is a wild card. So if I run the workflow here, we'll notice that it took all of these files, it read all of these files in. So about almost 12 files in here. I can actually, any time I read in the data to altrix, I can actually see what that data looks like just by clicking that little arrow or output node after it. Again, so we can see that again, we have transaction level data that we're going to aggregate to on the level later. So that's one way you can connect to data in altrix. Another way is we can connect directly to databases. So we do support almost all the major databases or anything that does have an ODBC connection. I have a SQL server that I'm going to connect to. So I can simply click that database right there. You'll notice that I can see all the tables in the database. If I already have a SQL query written, I can just go ahead and copy and paste it in here. But I actually want to get the full database. So we just go ahead and click on that. Now connection to a table that's on my database. So this is one more month of data that was sitting somewhere else. Maybe it had to have some kind of correction done to it. But now we have that connection established. So now that I have my connection established, I can start to build on my workflow, do my data preparation, what have you. So the first thing I want to do is union these data sets together. So I want to stack them on top of each other so we can get that kind of want together in one data stream. So if I go to the join category, select the union tool. You'll notice that it will automatically make a connection if you put it close to tool and I can manually make the connections by dragging it over. This tool is going to just stack the data set vertically. If I put a browse tool I can actually see the full data set as well. So I'm going to go ahead and run the workflow. You'll notice the record count that we have here 55,000 to 5,000 and now we have 60,000 records because coming out of the union because we stacked them vertically. If we look at the data itself you can see we've got, I'll just let you know with these little red arrows if the cell has trailing white space in it or needs to be cleansed. So the next thing I'm going to do is actually cleanse the data. So go to preparation I can actually use the data cleansing tool so go ahead and drag it onto the canvas. I want to cleanse all my fields except for my transaction date field. With this tool it's going to apply these functions to all those fields I've selected so I'm going to replace nulls. Anytime there's a null value in a string field I will replace it with a zero. I also want to take away any leading and trailing white space and any tabs or line breaks. So any kind of characters that I might not probably wouldn't want the data. So I'm going to go ahead and run the workflow again. And so if we actually click on the data you can see those little red triangles I had before are no longer there because they don't have any more trailing white space on this field and my null values have been replaced. Additionally if you notice that these were CSVs that came into here so I can actually look at the metadata itself since it's a CSV you'll notice that all of our field types are strings. We may not want that because we actually want some numeric fields we might want to do some math on them or some date fields we want to use some of the date time functions in Altrix. So we're going to use an auto field tool to have Altrix automatically type the fields for us. So go ahead and run the workflow. What you'll notice now when you look at the metadata is that you can actually see the new types in the data so you have integers, dates, and doubles so on and so forth. The next thing I'm going to do is apply a formula to the data. So if I use the formula tool I can go ahead and do so. The formula tool allows me to apply formulas to any of the fields or I can even create a new field by typing a new field name and that's what I'm going to do. So we're going to call a new field called month. I'm just going to have it be the string and what we're going to do with this field is we're going to format it. So if you looked at the data set before we have this kind of date in there but I want to actually create a new field that has the name of the month, January, February, March so I can aggregate on it later. And what is the month number? So within this formula tool there's a full functions library in here. So anything from doing if then statements, string functions if you have any kind of survey response data you want to look for a certain word using the contained function. What I want to do is use a date time function to format my data to create a new field and the function I'm going to use is the date time format. So I'm going to go ahead and double click that and you'll see a pop with this expression box and I simply have to change these parameters. So the date time field I'm going to use, maybe that transaction date field. And then there's a library within the formula tool that lets you know what you can tell it to get that out. So if I want to know transform that date which was 01, 02, 1994 into a one that just says fed, I can know that from the altrix dictionary that I can actually just put in percentage sign capital case B and that'll do that for me. I also want to do another field called month number to tell me what's the number of that date. I'm going to do that so I can sort on the data later before we put it into the time series model. So I can create a new field called month number which will also be a string. I'll use that same function, date time format. So choose that same field transaction date. But this time I actually want to see it as a number so I'm going to use the syntax to show me that. Again this is from the altrix dictionary. There you go. This is going to be lower case M. All right. Click on the canvas. Any read information went away. It means my workflow won't run if I run the workflow. I can see there's no new fields that I've created. So if you look at this formula right here, scroll to the end so you can see these new fields I've created. So one that's showing me the month number. And if I put up the browser to show you the full results we can look at the other months as well. So we've applied the formula now. The next thing I want to do is aggregate the data. So in altrix you can easily aggregate or create views of the data using this summarized tool. Apply that on the canvas. Find which month you wish to aggregate to. In this case it's going to be the month field. So I want the data to be turned from transactional level data. Daily data and aggregate to the month level and select the month field. Add the group by. I also want to see the month number in there so add a group by on that. Now everything I do, any metric out of now is going to be aggregated to this level. So if I want to know the total retail sales per month select this field. Retail, let me just expand that for you. This is the retail price for retransaction. I'm going to go ahead and add a sum. I also want to know the number of units sold per each month. I'm going to go ahead and add another sum. And in here I can actually also change the field name that's going to be presented. I'm going to change this to total sales. This is going to be total units. And it's going to place a browse tool after this. So again this browse tool is when we pop out the full data set instead of just seeing the preview at the node. So I'm going to go ahead and run the workflow. And what's neat about Altrix is you can see the record counts going across. So we had the 60,000 records, right? We aggregated to the month level. So now we only have 12. Let's look here. We have our 12 months. We have sales in each month and the number of units sold in each month. Okay. The next thing I'm going to do is sort the data before we throw it into the time series model that's going to predict for us basically what our projected sales should be per month on the next year out. So we'll sort it so that we have it in the correct month order. So I'm going to go ahead and sort tool right here. I'm going to sort on the month number. And we're going to sort in ascending. So I'm going to go on the workflow. And all you're going to notice is this tool is just restoring the data so now it's going to be in order January, February, March, what have you. So we see that here. Months, 12. Now that I have my data prepped I'm actually going to run a predictive model. I'm going to do a time series model. So you have the ability to remit ETS models. Again, what the time series forecasting model is going to do is going to build up our historical monthly sales that we have. Based on these historical monthly sales, try to figure out 12 months out. What are the, what can I expect when my sales to be per month, 12 months in the future. So go ahead and go to that tool. The sales field that we're going to target, the field we're going to target to forecast on is total sales and the data is monthly data. So really easy customization here. If you want to get into the weeds and you know how to be really good at customizing your model, you have further customization features in here as well. So this will actually produce, this tool will produce your report for us. What it also does is it's producing the model and then the forecast values to the model. We just use a TS forecast tool. Place that over the object output. And then put a browse tool there. And so the only configuration of this tool that we want to do, I'm going to, I can change the confidence intervals if I'd like. But I just want to say give me forecast, forecasting the future, 12 months out to the future. And let's go ahead and run the work flow. So if you want to take a look at the report model and how it did, I clicked this browse tool here. It's going to give us an interactive report showing us did it well, what was the information behind it. So you can see we've got some plots for ACF, PCF. We show here the trend of our monthly sales basically and what our forecast to do out into the future. The interactive map you get the values based on where our mouse goes. This tool produced the model which we can then use the forecast tool to provide the forecast for us. This is showing us for the next year out that second year for each month what we can expect the forecasted sales to be. So now I've finished building my model. I like what we've been able to do so far is to connect to two different data sources. So that's two CSV files. So we read in about 11 CSV files with one tool exactly at the same time. Altrix had the ability to union those for us. We read it in a SQL server table. So one of the sales tables so it was a SQL server. We unioned it, did our data preparation, our aggregation, built our predictive model. Now we're going to finally output this data set. So the whole goal of this was to prepare the data for power BI, so it can visualize in power BI. So Altrix has a connection to power BI cloud, so if you go to the connectors, you'll see that here. Also want to highlight as well, Azure ML text analytics, so if you're doing any kind of sentiment analysis text analytics on Azure, we actually have a connection to the API to actually pull those squares back to you. But since I want to publish this data to power BI, please publish the power BI tool. I'm going to set a data set name. You can call this retail underscore August. You can give the table name RA, create a new table on power BI, and we'll go ahead and run the workflow, and this is going to publish it to power BI for me. Generally, the first time you'll run this, it'll ask you to log into your power BI account. But since I already have, because I ran this before, it will probably automatically publish because I have the refresh token initiated. So now I've created this new table on power BI. I can actually go log into my counter in power BI, and we can see this one that I've just uploaded, retail August. So switch over to that data set and you can see the fields that are in there. I have the fields that are in there. I have the prep data set. I don't have to worry about doing the prep in here, and I could start to build out my visualization. Right? So I could start, you know, put out forecast, period, and I can start to build out my charts and my graphs. So that's kind of how the two work hand in hand together. You have all tricks as this tool that's basically doing, you know, all that data preparation, forecasting for you, so that when you open and sending that to power BI, so that when you're sending it to power BI, you're only worrying about really visualizing it. So that does that for the demo for me. Dan, happy to take any questions now or during Q&A? Yeah, no, thanks, Ali. Appreciate it. We'll move into Q&A because I know there are a few questions here that people have been posing. But before we do, just to conclude, so basically we've shown you how Mazda built a, we didn't show you their exact model, but similar to the predicted analytics model that they built up in forecast sales in any given month in territory. And then how they pushed that into power BI, the dashboard Miguel showed you isn't their exact dashboard, but it's similar. And then that's what helped them be able to gain insights. So using those two tools, they were able to again, reduce the complexity and time spent on data preparation tests, eliminate any delays in getting to that information quickly. They leveraged predicted analytics with no coding. And then as a result, they were able to generate more powerful visualizations to get to those insights better. That way they weren't anecdotes. So with that said, let's move on to Q&A. We do have a few questions here. And I'm going to start with you first, Ali, just since you are the most recent person. I have here an attendee that asks, does altrix allow you to manually type case as opposed to the autofield tool? Yeah, definitely. And if you can still see my screen you have, so what I should do with the autofield tool, if you want to manually do it, there's a tool called select. Let's just go and drop it in here. And select tool, you can see I can see my fields in here if I need to change anything from the screen here manually to a double. But I want this to be an integer. You do have that ability to do so. Okay. Awesome. And I have someone here. Can you review the creation of the model step again? If they're referring to the time series modeling tools, so you notice that, so this is doing an ETS model, so you have two remaining ETS. I'm choosing which is the field I want to forecast into the future. So I selected that field. It's asking me what level is my data, what is the time frequency of my data? And it's monthly, so monthly was already selected. So that's basically all the configuration I did there. For the actual forecast, for the data that shows me what the forecast is, I use the TS forecast tool. And all I showed here is basically it's how many periods into the future do you want to forecast. And I set that to 12. Great. Okay. Great. And then, just because there's a lot of questions coming on this, is there any chance that one can run a partial workflow? So what you can do in Altrix is you can actually disable pieces from running. So I can put a tool container around a piece and have it not run. So if I had a separate branch, maybe I didn't want to run this every time because I wanted to do a separate branch here. I can put a tool container around a piece and have it not run. So if I had a separate branch, maybe I didn't want to run this every time because I wanted to do a separate branch here. I can put a tool container around that to disable it. But currently you cannot only run just a piece of it. That is changing with our next release of Altrix. So currently you can't, but you will be able to. But I do have the ability right now to disable pieces from running. By putting it in the container tool. By putting it in the container tool. That's cool. Okay. So this next question actually ties into both products. So someone asked as you're working with data and you're building on your tools and getting all that data prepped. Is there a way to automate the publishing to Power BI from Altrix? And if so, what frequencies are available? And then Miguel, to follow up that, I'll ask, while he's not answering that, I'll ask you a follow up question, Miguel. Yep. So yeah, you do have the ability to schedule where it was to run at a certain time frequency. You can pretty much set any frequency you want. So directly from here, I can either schedule this to my desktop or if I have an altrix server, schedule it directly to the altrix server. Let me just show you that scheduling piece real quick. So save my workflow. You have a scheduling console where I can schedule it at once, any minutes, days, months, a custom schedule. So maybe run it to the third Friday every month. So you have that ability to do so directly in altrix. And then so when you schedule that and obviously the ending piece of the workflow is to push to Power BI, that will also automatically update the data in Power BI. That is correct. Yep. Cool. So that you could ask. Yep. Go ahead. Okay. So Miguel, to follow up with that, once the data is updated in Power BI and you go into your Power BI account and it's there where you need to go, to follow up with that, can one share the data, you know, with others, the dashboard, not the data, sorry, the dashboard with the visualizations in Power BI can one share that with others. Exactly. Yes. And actually great point because I don't know how I missed that from the demo itself. Ali, you can give me the screen back. I can actually show where the button is. It's very easy to find back here that we'll put it. Here we go. So back here on the dashboard that we were looking at before, you're going to see on the top right here, share button. It's as easy as clicking on share button. You can get access to either people within your organization so you're seeing the main or people outside your organization. The other requirement is that they all need a Power BI account. That's it. You're going to include an optional message too and then some settings that allow you to either allow you to allow the people you're sharing the dashboard with to share it and if you want to send an email notification. And once you are part of that sharing of that shared dashboard, you can go back here. You can actually see notifications so these are all the dashboards that are basically being shared with me. Is there any changes? Alerts? If the data changes, you're going to get also a notification here or if you choose to an email into your inbox. Great. Now, staying on that screen Miguel and staying with that dashboard, another question. For the Power BI sales dashboard, the one that you're showing right now, is that a pre-built dashboard that can be shared with users so that we can utilize it for our own purposes or was that built from scratch? This was built from scratch. You have several options to that. So this particular dashboard brings, going back to what I explained before, all the different components that you have here on Power BI, the dashboard is something that you build every single time you have reports. So once you have a report here, like this one in particular, you choose what you send to the dashboard. That's the one that actually let me make this bigger. You are the one that chooses what goes into the dashboard. So that one in particular, it cannot be saved as a template. The reports themselves, depending on how you create them, this can be saved as an actual Power BI desktop file. I'm not sure you can actually download this. I think that's on a roadmap, but I'm not sure it's out yet. And the particular case of outbreaks where you connect directly to a cloud service, it's going to be very hard for you to replicate this as a template. What we actually provide people with, if you go to our community website, we have all these dashboards, actually the Power BI website and the community website, you have all these Power BI desktop files that you can download and replicate as much as you can. You can see the structure of the model, if you want to replicate that on outbreaks, and then what are the actual visuals that we're using. So if I would have to give recommendations to someone that let's say you want to replicate exactly what we have here, the easiest way to go would be to build a model with outbreaks first and then kind of go one-on-one on the different, the actual data that you have and replicating those visualizations. Power BI desktop allows you to say template, so once you do that and you export it to Power BI desktop, then you would be able to replicate that. But it's very specific to the data model and the data structure you have. So even if we make this available, which we actually did, based on another webinar we did a couple of months ago, it's going to be very hard for you to just take that and replace the data and come up with the exact thing that's worked for my template. Okay, great. Sticking on that report, so two questions here actually related just to this screen. We had one attendee ask, does Power BI tool allow you to schedule reports to be distributed? And the other one is in regards to reports, where did it go? It says just to be clear with Power BI, you start with the reports and then create the dashboard or is that not true? Correct. So I should have been more clear. Let me answer the second one first. So if you remember and actually analyze demo two, once you have the data set which is what we had here, this is where you create the report. So you start drag and dropping like different types of KPIs and measures and we were doing here before with country. So let's say this is my very complete dashboard which has a bar chart of sales in countries. You state this. That's going to be a report that's going to be saved here. And once you have that complete report, like the one that we showed before, then you start bringing back to the actual dashboard. So let's say you don't want to share all this information, so you just pick and choose the one that you actually want to send to a dashboard like this one and this one. That's where you actually get that over level view which is the one that you actually share. So when you share something with a colleague, this is the first thing that they're going to see. They can interact with it in different ways like in Q&A, click on any tile that will take them to the actual report. But the process where you build this dashboard is you need the data set first which is what we show with outrage. You build a report which is what we showed here on the browser. Once you have the report, you select the specific visuals that you want on the dashboard. Even images which is the case that we have here, these are just like what we show. We're going to show people on the dashboard what's information being displayed. That's where you actually get this end product dashboard that we're seeing here. And then can you remind me about the first question? Oh, schedule. The reports, can they themselves be shared? Yes. No, no, you, sorry, if they can be shared, they can be scheduled. Is that what you're talking about? Sorry. So does the Power BI to allow you to schedule reports to be distributed? Oh, okay, perfect. So the idea in Power BI is that whenever you get to the system, these web access, the mobile apps, any access points for Power BI, even the emails that you get telling you that data was updated or say, I'll see the threshold, all of that data is going to be up to date. So there's actually no need for you to actually keep reminding people that the data is updated or that you updated something. Every time they go to Power BI, they're going to get the latest. So it's going to depend, again, in the specific case that we showed on the webinar today, if you schedule updates from outtracks as long as those are happening and successfully being pushing to the cloud, every time somebody comes into this dashboard, they're going to get the latest data. So by default, there's no way of just reminding people, remember that that's where it's here. The way that's going to show is going to be how you set up the emails and the updates and the alerts for people to know if something is being shared with them, if data was updated, or if you actually had an alert set up on one specific KPI, that's the threshold, let's say, sales fell under $2 million per month. You get an email in your inbox, you click on the email, they take you to this dashboard, and you get the latest view of the data here. Okay, great. So I just have a couple more questions, and then Shannon, I'll pass it back to you. Two probably, one for Miguel. Miguel, can a PDF be created of the dashboard itself? Yes, of the dashboard itself. That's a great question. Actually, the print option that we have is on reports. So we'll go back to the report here. You can see here on the file that you can print the report. It's not PDF by default. You can use any PDF writer to just export this into an actual PDF. But you can print it, capture it, or export it to PDF any way you want with this option. Okay, great. And then Ali, back to the Altrix workflow. First question, can the workflow be initiated via an API call? Sure, yes. So we do, there is an Altrix API. So with the Altrix server, you can actually call an Altrix workflow to run and return data back to something like a web service. Okay. And then second question, can the Altrix workflow end points be published as services such as RESTful services, SOAP services, et cetera? Yeah. So kind of, I think that's very related to that, the first question. So you can have via an API interface, you can actually pass data. So the input end point, you can pass data to an Altrix workflow and then have the other end point, the output return via the API to whatever is calling it. Okay. Awesome. Great. Well, thanks everyone for some of your questions. Hopefully this was helpful. Shannon, I'll pass it back to you. Thank you all for this great presentation and demonstration. And just to answer another question that often comes up, just a reminder, I will be sending out a follow-up email within two business days. So for this webinar by Ender Day Thursday with links to the slides, the recording of the session and anything else requested throughout. And thanks to our attendees for being so involved in everything that we do and asking such great questions. And then of course we want to give thanks and shout out to Altrix for sponsoring today's webinar. I hope everyone has a great day.