DataSelf for Manufacturing with Natural Language

Let me get started. So hello everyone. Thanks for joining us. My name is Johnny Girardi. I'm Data Sales founder and CEO. First of all, I hope you're doing well staying safe. Insane. Which has been my biggest challenge, thank God. Anyhow, we have a lot of good content to cover. Today I will be talking about data cell for manufacturing companies using analytics, natural language, machine learning and artificial intelligence. I'd like to first of all introduce you to Clark Wallister. He's going to bring some interesting perspectives into this webinar. So Clark, would you please introduce yourself and go over your first ideas. Okey doke. Hello, my name is Video by the way.

You can see my rustic office. My name is Clark Walliser. I, I've worked as a for the last 30 years. A lot of my career was as a Sage reseller consultant. In that sense I focused on manufacturing from the reseller side. In later years I went to work for an end user using manufacturing. So I got the experience of implementing manufacturing on Sage and Epicor in there. Then I went to another place as an end user using Dynamics 365 was involved in the manufacturing and development there. I've created a fairly broad base of manufacturing from both the consulting side and the end user side. The latest phase of that is when I joined Dataself. It's been a year and a half ago and now we're moving into data data cell for the manufacturing how to build on what the ERP systems can give you. But to get quicker, more effective results for the data, you need to do the manufacturer to do your manufacturing effectively. Does that kind of give you an idea there, Johnny?

Yep, Perfect. How about you talk about the first use case you're going to be covering?

Oh, you want me to cover that right now?

Yep.

Okay. Yes, as Johnny mentioned we, we have a few use cases of where we are using data self right now. This isn't vaporware, this is actually being used by customers. The first one is very simple. The, the next two will become, you know, more exotic and hopefully interesting. The first one, we, we have one client that they were able to, you know, they had an ERP system and they were able to pull the basic information out in reports etc etc and they were trying to build dashboards but it was the performance wasn't there and to pull the data out in a timely manner and then the work to get all the right connections and everything was overwhelming them. They wanted to do the work, not spend all their time developing a report on how to do, on how to do the work with Dataself we've implemented the manufacturing features for their system and so now they have that data from the ERP system. It's been massaged and cleaned up so it's all in one data set for them in the data self data warehouse. And then when they run a report it's just from one data table. There's no linking involved and they don't have to pull the original ERP system, it's just right there. The, the data self data set is refreshed every night so they can just run a report in a matter of moments instead of hours and, and then they can lay out the data however they want. They don't have to do calculations for every single report. They had tried things in crystal reports or reports reporting services, other tools. Before they were simply report writers. They just pulled the ERP data source and it was live every time, very slow. The difference with this is we built the data self data warehouse and they can control what data goes into that, create it all into a flat file, simple for that, with terminology they understand they can, they can refine that data source to be exactly how they do business, not just how the ERP system is designed. And now they can get output in a matter of moments. So to me that had value. I was really. They aren't doing anything exciting right yet right now, but they're able to get the data when they need it.

Yep. Clark, very good input. I remember this client they mentioned, you know, some of the challenges that you're talking about, you know, more in detail and they have been one of our happy clients where even though the data self setup was not vanilla, you know, we put some effort to help them get to the point where the system is really performing quickly but they invest their time into learning how to maintain data self and now they're very self sufficient. You know, they have been slicing dice in their manufacturing data and they rarely call us for support and that's awesome, you know, giving the power to the user so they can become more self sufficient. I'm going to go through some of the marketing sales perspective of this webinar and throughout this whole webinar Clark will come back again twice at least to give the other use cases that will kind of be part of this content. So Clark, thanks for the first one. So I'm going to talk about the very high level manufacturing companies. Some of the good news is they have a lot of valuable data to help them become more efficient, more productive, more profitable. That's the good news in very, very Commonly, at least in our client base. These are the systems, the ERP MRP systems that our clients use to handle their manufacturing. We are broadening the number of ERPs, but currently these are the ones we cover. Now these organizations, they use the systems, one of them, maybe more than one, but you know, at least one of them. But they have more data silos feeding into the whole business. You know, maybe CRM from different systems like Salesforce, Excel, usually all over the place with a lot of information being stored only in Excel. Some people have HubSpot, you know, E commerce. I mean there's a lot of different data silos that organizations use to manage their business. Customers, marketing, sales, inventory, payroll, you name it. There is a lot of data silos. That's the good news. The bad news is, well we call this is a data house divided. The data is there, it's right there. But if you need to take information out of it to, to make a better decision making process, it's not simple. Typically management complains that they're flying blind. Information is right there, but they don't know what's going on. When they get to know, it's too late. So flying blind is a complaint we hear quite often from decision makers out of this data house divided. And ultimately a lot of the times it becomes what some people call Excel. Hell, people export data out of the systems into something else. Excel is very popular Crystal reporting services. No matter what people do, a lot of them, they get the data into a different format. You know, again Excel is very popular and sometimes the data is outdated, the Excel is broken. Nobody knows Excel crashes or the reporting tool. So it can get very convoluted to control the data from all of these different systems. And sometimes the answer to some questions is between two different data silos and how to get the information easily and quickly to management. That's complicated. Overall the main reason people look for Data self, one of the main reason is labor intensive for management to wait for the information for the people supporting the decision making process. It takes a lot of effort, a lot of, you know, it's very error prone process and expensive and slow. How do we help companies going through these challenges Like Clark mentioned, you know, we do use a data warehouse framework at the heart of the company. Data Self is a data warehouse company and what we do is we bring your data silos into a data warehouse, we clean the data, we optimize it, we pre link things, we use SQL Server as a platform and we use our own extraction transformation loading tool to make deployment and maintenance of data warehouse. Something simpler, easier, faster, more optimized. Then we plug very popular and successful user interface tools like Tableau, Power BI and Excel. And out of the box we typically bring thousands of reports and dashboards that are ready today, ready to go on day one for manufacturing companies. Usually the manufacturing components that I'm going to show very quickly in a minute, they're not out of the box because usually manufacturing requires some tweaks. But there's a lot of, you know, out of the box reports and dashboards in sales, in inventory receivables, payables, that is, have proven to be very valuable for many of these manufacturing companies. So that's how we help as a platform. We've been doing this for 15 years. So the solution has evolved with a lot of fine tuning, helping the system stable in the long run. Even if you upgrade your ERP and CRM, typically it's easy to adapt our whole data warehouse analytics platform as you change and as you grow. Having said that, Clark, how about you come with the explain the use case number two and don't forget your web camera.

I even remembered to unmute myself.

That's helpful.

Yes, sometimes that's helpful. Our next use case is somebody that they are driven by by sales order and the sales order and projections. Sales projections. And their big driving force was to get find out what raw materials and sub assemblies needed to be created at which times. And they wanted to be able to eliminate shortages and overstock of material. They wanted to be as justice in time as possible. And so their ERP system could do that. I mean, they did have a theoretical mrp, you know, an MRP function, but the data in the system was not always the way the erp. The MRP function needed it. I mean, their POS were not all necessarily clean, their sales orders were not always clean, et cetera. So they weren't able to really make good use of that with the manufacturing features of dataself. This gave them the ability to customize how they identify sales orders as a supply, purchase orders as sales orders as demand, purchase orders as supply, and then also work in progress, work in process for potential supply there. Each one of those cases had exceptions where they did not want to include some records. With Dataself, we can pull from all those data sources, put them into one flat file data set, and then we can include all those filters, as advanced as they need to be, as complex as they need to be to identify what qualifies and what does not, qual does not qualify in that. Then the buyer actually he can just open up his ui. In this case he's using Tableau and he has his pre made reports, but he also does a lot of ad hoc queries and the data is just right there. He's got a list of this, 50 fields that he needs, he doesn't have to link tables and he can redesign and relook at the data as he needs it without having to recreate it all from scratch. If he was using I think crystal reports before and if you wanted to take a slightly different look at things, you'd have to either modify a copy of an existing report or relink everything from scratch. And all that this eliminated all those issues.

All right, thank you Clark. Moving along. So now I'm going to be showing the Data Self demo, touching some of the points that some of our clients have used. So my idea is to show several of the Data Self manufacturing framework that usually when we go to manufacturing company they are set up again a lot of the manufacturing framework out of the box. We have pieces that usually can quickly be plugged to the system. But unlike a lot of the out of the box solution that itself that plugs and play on day one, manufacturing usually requires some additional setups, sometimes quite a bit of setup. But once it's done, what I'm going to be showing you is the result of once the system has been set up and give a little bit of a story based on some of these use cases that Clark has covered. Let me switch gears and let me reload my page here. So the very first concept is very high level is the data house united approach of Data Self. The idea is again companies might have all sort of different data silos in their business. In reporting usually every tool has its own reporting. You have to have the log into that system or someone has to export the data out into Excel or something to build reports and that's across the data silos. With Data Self data Warehouse all the data can be together and when you log to your Data Self portal which can be embedded in your own application, this could be embedded in your Acumatica, Your Sage, your NetSuite, Microsoft Dynamic system so people don't have to go outside to see Data Self. It can be inside of the framework, but when they log in using user credentials, the system can control what data sets I have access to. Right now I will be let's say the CEO or the cfo. So I have access to my ERP data, my CRM, E Commerce, Google Analytics, more ERP payroll, anything would be all Inside of Data Self you just point and click and easily you can go to different data silos and you can even blend them together. You can have a single dashboard that will have manufacturing data, P and L, Google Analytics, data, payroll information, E commerce, all the data again can be brought together easily and quickly. So management can be keeping a close look at the operations and finding opportunities and threats now, not days, weeks or months later. So that's the approach of the data house united with Data Self. Now most manufacturing companies, they also have a strong distribution business as part of the whole operation. Today I'm not going to be covering the distribution analytics we provide. If you want to talk about and learn more about the sales side and all the things we can help with the distribution side, you can go to our Data Self website, go to dataself.com, go to industries and select distribution. Last two days ago we recorded a webinar focusing specifically on how we help, especially sales analysis, some inventory as well as receivables and financials. So feel free to explore this and learn this page to learn more about the distribution side of the business. What I'm going to be focusing here is mostly this column here which is inventory management, super, super important for manufacturing and more specifically around the manufacturing components. So to get started, you know, manufacturing companies, you have your open sales orders, you can analyze them, you know what's, what's. If you have missed any shipments, you can see what products you have on hand to fulfill your orders. So all of this section here is what's ready to be shipped, how long it's taking to be shipped. Is something falling behind, do you have slow moving items? So this section, you can analyze those metrics, but again that's not manufacturing that sale. So I'm not going to be covering that more than what I just told you. Let's focus on manufacturing. The point is you cannot ship some products because you need to manufacture them. So typically the first step is inventory planning. So let me show one example that we have customized or we have deployed this kind of template to analyze inventory planning needs. So what I would have here on the far left would be my finished good needs. Which items that I have in my, you know, finished good list I have here for the period, the current week or month or day. So these could be, you know, different time periods. In this example we're looking at week and this is the current week. The first row for each product is the how many items you have on hand today. And then you can see how many items are going to be coming to my own hand from the work order. So in this example I have nothing in my work order, how many I'm supposed to be shipping by week. So these are open sales orders. And what is the sales projection to be shipped by week? How do we calculate sales projection? We usually ask the client, how do you calculate sales projection today? Sometimes we just have to replicate the process. Sometimes we can come up with ideas to enhance your sales projection calculation. And sometimes we're not the right good fit to make sales projections. So usually we look at the scenario, usually we can improve it. Sometimes you need some dedicated sales projection tool to fulfill that particular line. The good news is once this line is calculated by data cell for some application or some human we plugged over here. And then the tool calculates the on hand, the quantity on hand projected by week. And when it's red, mean you are below the minimum quantity on hand. And this is the excess short amount. And again when you have a red column, it means in that week you're going to be running out of inventory on hand. So the idea is you need to produce more so you can fulfill orders or buy more in case you can also, you know, source some of your finished goods from third parties. But this is the kind of framework that we use to help clients do a good solid inventory planning. You can have, you know, here models of changing the projection and looking at the ABC product category and warehouses, whatnot. So this is inventory planning. So inventory planning. Once you even you do inventory planning, you realize I need to produce this in this time frame. Well now you need to see if you have your bill of materials to fulfill those orders. So people will go to the bill of materials data cell framework. And this is an example of bill of materials that give a perspective for some business can be helpful. By the way, what I'm showing you is several different sample databases. So don't try to make sense if the data from this slide is the same data from my prior slide. Currently they are not necessarily. This is just a sample database. I'm just covering the flow of the process and how these reports and dashboards help with each of these steps of the manufacturing decision making process. So the bill of materials, what I have on the far left would be these are the items that I need to produce. This is the current quantity available in my inventory. And the inventory can be in warehouses, can be on water if you're shipping things from Asia, China and whatnot. But anyhow, this is considered what is available in my what is available today. Some companies the available means you have here ready to ship. Some companies they consider available something that is already in transit in manufacturing. Sometimes every single column of our reports is a discussion, is a needs analysis and then eventually hard coding the whole data self system to automate the data manipulation of that data point. So you don't have to do things in Excel, you don't have to do any exceptions. The computer will be calculating the data and quickly and easily give you the results into a tableau, Power, BI or Excel sheet. Anyhow, so here I have my quantity available. These are the finished good quantity requested for my work order production list. And then here it tells if you are short on items and every time that you're short you have a red, you know, this item here you cannot fulfill the whole order on this side. Here we actually have the bomb exploded into its own bill of materials. If you have multi level bill of materials, we absolutely can help you a lot out of the box. We have the templates pre configured to go from level one to level five. So you just have to tweak that framework. But if you have more than five levels, 7, 8, 10, 15, we can absolutely extrapolate the framework to go as far down on your bill of materials as necessary. So you can see the finished good were the level 1 components, level 2 components, 3, 4, quantity requested per BOM, how it's available and then for every level of available materials, what the quantities are at the bottom. We pretty much what we do is we take the raw materials by BOM and here we have all raw materials to fulfill whatever you have here. So a big way to look at all of these data points. And this is highly interactive. If I select a bill of materials here, a finished good, it shows me only the components across all the data sets. So very quickly you can point and click and see if you can do it or if something is missing or if you're managing your raw materials and you realize that this item here is not going to arrive on time, there's a problem. You click this raw material and then it shows you which of these finished goods are being affected and then you can maybe manage it. Hey client, sorry, it's going to be late. Or can we replace this item with something else? Because our raw materials are not going to come back in time. So very interactive way of looking at building bill of materials. Once you come up with the bill of materials and you finished what you what you want to produce, then the idea is you go to hey, maybe you need to buy some of these items, right? In order to produce you might be short in inventory. Then you can go to the purchase order and analyze which items you need to place an order to be sure you can feed them into your bill of materials. And this again is an example of an out of the box open purchase order dashboard. So the red are the items that are late. If I click on the red, it shows me every single item by vendor, by product and all the details of being late. So very interactive ways to analyze vendors, products and whatnot. Before I jump into the next section, which would be work order and so forth, I'm going to call Clark to join us again to give us a perspective into the third and final use case.

Thank you. I'll just adjust my camera there. What Johnny has just been showing pertains a lot to this third use case. This client. I can't use their client names for obvious reasons, but these are actually live clients we're working with, trust me. But as Johnny was showing you the examples of those Data Self views, the dashboards there, you can create dashboards in your ERP system. You can create all of those. Why do I need Data Self? Before I worked for Data Self, that's the question I kept asking Johnny and he says, go ahead, do it. I don't have time. That's a lot of work. That's, I mean, man, that's all that together, that's going to be a huge investment. He's going, yeah, or you can just do it right out of the box here. And that helped me become a believer in Data Self. While you can do all of that in the RP system or in other systems, the alternative of having it done for you out of the box, you get results immediately and then if there's issues, you have a whole team of people to back it up and to keep things updated, et cetera. There's a lot of value in that. But now with this third use case scenario, very similar to the second one, they were concerned primarily with raw material and sub assembly, supply and demand. Because you know, my kingdom for one screw that I ran out of on the production line. So my $10 million finished, good, cannot get shipped. We don't want that to happen. However, their sales order activity is so volatile they have an E commerce. And so they might have 1,000 work orders. I mean sales orders come in in an hour. So they are, with the Data Self they're able to take the supply sides that are not as volatile and refresh those nightly. But on the sales order demand side, they refresh that every hour of the day. So they can get a current view of what's going on. You're ending up with one flat file of everything that they need. So it's fast to report on, but we can refresh that flat file on a recurring basis. And that really made the difference for them because they were trying to do this in crystal reports before. The crystal report would take eight to ten hours to run. By that time the data is stale, there's nothing, it's nothing even close to what they really need. So this provided them the uniformity of data consolidated in one flat file and the speed. And then they have the ability to take the canned reports, data views, start with those and modify them as needed, or create copies of them to tweak it for their ad hoc scenarios without having to rebuild the entire structure. I think that's pretty much it right there.

Yep. And just to complement your story, Clark, this client, as Clark was mentioning, many of these crystal reports were taking several hours to run. So they were having a hard time managing their manufacturing processes. And it was getting worse. When they started with this process many years ago, it was not too bad. You know, the crystal reports was running fairly quickly, but as the data accumulated, eventually it become unbearable. Make a long story short, now that the data itself has been deployed, those reports were taking several hours to run. Run under 10 seconds.

And I'm back with. To make the short story a little longer, a couple of things I wanted to point out that are significant to me. This client, they used sales orders in many different ways that the ERP system was not necessarily intended for. And so they wanted to be able to control which sales orders were included in the demand part of the calculation. And so they with data self we can build that logic right into the data self data set without impacting the ERP system at all. Same thing on quantity on hand for supply, they have the multi bin inventory and they had specific bins that were qualified for supply of raw materials and sub assemblies and ones they wanted to exclude. Again, with the data self we can build those filters and that logic in without having to do anything at the ERP level. So I just wanted to point that out. That to me is a big, a big feature that enables me to adapt to my specific conditions.

Good point, Clark. Yeah. And one thing also that we want to make clear, we are not replace data step is not an MRP system. We are a reporting system that sometimes we can augment your ERP MRP system when reporting out of it becomes slow, complex and crashing your system. So again, we're not an mrp. But we can help with several steps of an MRP process that sometimes the reporting tools available in the systems are underpowered and overly complex to get the job done. All right, I'm going to finish my product presentation. So we stopped at the Noblo materials purchase order. So work order is usually what Clark was talking. Once we have your work order into a flat table, then slice and dice in your work order data by different steps. The details, the production centers. This part is usually not the, you know, for data geeks like me, this is the boring part. Even though it's what people really live and breathe in the manufacturing plant floor. But work order is fairly straightforward. How you analyze all the possible steps of the manufacturing process. Many companies super conscious about, you know, looking what they have planned, plan to produce and what the variances the items that were, you know, they lost in the production, scratched production, how much cost was involved in the producing the goods. Many of our clients actually, they're still struggling to capture the actual data of hours invested. But if they managed to bring the data, you know, by, you know, somehow in the, in the season, we can track the time invested into each unit produced, then we can give a pretty good visibility of the cost not only from a raw material standpoint, but you know, from labor and maybe utilities if you want to include that in your cost analysis. The last dashboard I'm going to be showing you is a custom dashboard that we've been using to show high level picture that usually upper management would be looking at. So upper management, they might not want to look all those details and deal of materials and whatnot, but they want to see big KPI perspectives. So like in this example, you're looking at your items produced, units produced and how much percent of them were lost. The bars are the items produced. These bubbles are the percent loss for each one of these items. Then you can see average hours produced by product and also by operators. So you can see, you know, average time and why some people are more efficient than others. You can see what you have to ship in the next days. And if the fact is something you're trying to keep, you know, close eye and try to reduce them, you can easily plot the fact, you know, reasons locations so you can kind of, you know, quickly see problems and address them before it's too late. Again, these dashboards are highly interactive. You know, if I go over here and select these two items, when I select them, it changes the whole dashboard to show me people working in only those two items. What I need to ship the next days and what the defects are associated those products. So again, highly interactive way for upper management to analyze manufacturing big level KPIs. All right, the last thing I want to talk about is how artificial intelligence, machine learning and natural language helps the manufacturing process in our journey towards that direction. So the first thing I want to show very quickly is actually the natural language query built into these tools like Tableau, Power, Bi and even Excel. You can plug these tools to clean data in the data warehouse and then allow decision makers to ask questions to the computer to Data Self. And the answer to the question is a report built instantly accurately. Right now I'm connected to sales data. So if I come over here and I ask the computer show me sales by salesperson in 2020. So I just ask a question. The tool give me hits. Like when you're searching the Internet, you see the hits, you find the one that matches your question. And when you click, what happens is the solution will build in this case a chart. So by salesperson in 2020 is showing sales in a descending order. This is not magic, this is technology, artificial intelligence and machine learning to answer questions placed by humans very quickly, show two more examples. If you want to break down this by month, you just type by month. Again the tool give me the hits. I select the right one. And now what it did in 2020 by month is showing me sales by salesperson where the amounts are showing us a line chart. So quick question, quick answer. And maybe I don't like charts. I want to see a kind of grid or Excel kind of look and feel output. Sure. I can select tax table. Now this report becomes a text table that I can see by salesperson all the details. And if I want to export to Excel, let's say I want to export to Excel my January numbers. I can just select those numbers and tell the tool to view the data. Meaning show me only this data set. Finally give me all the details. So if this was manufacturing like work orders or bill of materials, this will give you into an Excel file every single detail behind behind those data points that have been stored in the data warehouse. So quick and easy ways to get the data out of the data cell system. This is the machine learning in artificial intelligence framework at the user level standpoint, so people can quickly get data out of it. But Data Self is also building machine learning and artificial intelligence on the back end of Data Self with two main purposes. One is to help you more quickly do the data consolidation. The Datahouse United we're building frameworks to facilitate the People involved in data consolidation. Second is also machine learning process to look at how to more efficiently extract your data. What's affecting maybe issues in your extraction process is your operating system, is your Internet connection, is your IP provider. Our goal is to help our clients to spend as much energy as possible in analyzing your business in your data. While the BI framework, the technology, the tools to help you get that job done become more easily automated leveraging this great machine learning and artificial intelligence technology. All right, so that was the main product presentation. I hope you have enjoyed that session. Let me go back to my PowerPoint slide to wrap it up. Data Self pricing starts at 299 per month. This pricing is usually well let me show you how you can find the pricing for your own specific case. So if you go to dataself.com so let me go back to dataself.com, pick the main product that you work with, let's say Acumatica or Sage or Dynamics. I'm going to click Sage for now. I know many of you are watching today are from Sage. Then you select your Sage product. I'm going to choose Sage 100. Now when you get to this page of your own ERP product, you're going to see these two options, Advanced analytics and Enterprise Analytics. Advanced analytics, quick and easy, deploys in about 30 minutes, very inexpensive to get started and we can tackle a lot of reporting needs for manufacturing. Specifically what I showed you today. We are still working on the mapping of the manufacturing templates and we don't know how far they will go but sales, inventory, purchasing, sales orders, receivables, payables, financials, Google Analytics, a lot of things can be tackled with advanced analytics. Currently all our manufacturing clients doing manufacturing Data self they use the enterprise analytics which starts at 1499. Anyhow, this page that you landed, if you scroll down you're going to see more details about this framework including some screenshots if you want to see what comes available out of the box for advanced analytics. If you scroll down further, you're also going to see the pricing sheet of Data Self Advanced analytics and you scroll down some more information including the architecture and if you want to deploy it again. Deploying advanced analytics is typically done in half an hour is super quick. Enterprise analytics is usually one week plus usually takes several days, sometimes weeks to deploy Enterprise analytics. So anyhow please check with your reseller or with Data Self if you want to learn more details, see how things work. If you decide to go with the advanced analytics solution, we even offer a 14 day free trial. It's limited availability, so if you're interested, check with your reseller or contact us to learn more details or just run with it in our website. If you click Deploy now you can fill out a form, put the credit card information and we'll get the ball rolling. Having said that, I'm going to open the Q and A session and before I open the Q and A session, just to remind you what we do, we do Datahouse United easy access to key performance indicators through reports and dashboards. A big value for manufacturing companies is the ability to automate complex and labor intensive manufacturing reporting. If you have complex, labor intensive and slow to perform reports out of human manufacturing, talk with us. Most likely we can help you a lot. We are on an amazing journey leveraging natural language, artificial intelligence, machine learning, a lot of things already built in our framework. In the future of the benefits of this moving forward is going to be even more amazing. Deploys in minutes and it starts at 299 per month. So if you have questions please use the Zoom panel. Let me ask some questions. There's a question here from Carl. How long does it take to learn how to customize Data self? Carl, thanks for your question. It depends, what do you want to learn? Let me just go back here quickly to my system architecture to answer this question. So Data Self has this kind of, you know, these three different frameworks. There is the data warehouse framework. Customizing this framework you need to be a technical person, you need to understand tables, fields, databases, and you should preferably know SQL Server, I'm sorry SQL statements. If you have those skill sets. It might take 8 to 24 hours on average to learn how to maintain the data warehouse. At least the 80% of the customizations you need to do on the back end of the data warehouse. If you know tables, links and SQL statements, typically in 8 to 24 hours you'll become fairly self sufficient in customizing the back end. This part of the tool which is the analytics and today I showed you tableau. Many users have learned how to use this with zero training. So if you're a self taught kind of person, there's a lot of materials out there to consume. Videos, blog posts tell how to do things. So many of our users, me included, never sat through a formal training of this tool. And I'm a tableau wizard and the same thing applies to Power BI in Excel. But on average people are busy, they don't have time to keep on reading a lot of things. So on average we recommend 8 hours of training for the average user to become fairly self sufficient in customizing reports and dashboards. The last skill set is someone that can look at the whole framework, talk with the users and being able to talk with people who maintain the solution to make the magic happens. We have several clients that they became completely trained and they do everything by themselves. We have clients that completely rely on us or on our reseller to do customizations. In the data cell framework, the average client is a mix. You know, they take over some of the customization that they can. They have bandits to consume to take over. In some of the more complex customizations they rely on their var or in data self. Recently one of our clients, an X3 client, they told us that when they decided to leverage data warehousing in tableau for their X3 reporting, they had two options. Start from scratch. They were saying I have to build my house, you know, the data house as I mentioned, versus Data Self. The house came pretty much ready. They just had to customize it and tweak it. They said by using data self they eliminating the need to hire three more people to build their data house. You know, not the data warehouse, the data house, the whole framework. So we dramatically accelerate the time that it takes to build, to deploy and to maintain a robust scalable enterprise level solution. Carl, Sorry that my answer was very was very lengthy, but I hope I have touched many of the points of your question. Looks like. Okay, maybe there's another question here. Oh Seth, the distribution webinar was fantastic. Thanks for the feedback. I hope you also liked the manufacturing webinar today. Anyhow, looks like we're out of questions, so I'm going to move on to the last slide of my PowerPoint. In case you want to contact us here in my email, our phone number website. I really appreciate the time that you joined us today. Again, stay safe, well and sane. Hopefully, like me, your main problem is staying sane. Thank God, because some other people are having bigger issues. Thanks for joining us, having a great day.