Migrate to cloud-native business analytics with Amazon QuickSight

"Uh welcome to migrating to cloud native business and analytics with amazon quicksight. Uh my name is Sean Boon. I've been with AWS for almost a year now and I'm a product manager on the quicksight team.

I'm gonna lead off the presentation today and joining me is my colleague Rahul, who is going to demonstrate some new exciting quicksight capabilities that we launched this week that will give you more reasons to move your analytics to the cloud.

And we are also very honored today to have a customer with us, Max from Siemens who's going to demonstrate his team's journey using quicksight.

So by way of agenda, I'm going to kick things off describing how we're going to put data to work in the enterprise and what that looks like. Talk about some of the challenges with respect to what we see in terms of fragmentation in BI. And then I'm going to describe our solution, which is Amazon Quicksight, which provides a unified BI service at hyperscale.

We'll then talk through some use cases for Quicksight as well as some new functionality and then, Rahul will do a demo for us on a new capability, a deep dive into paginated reporting. And then Max will join us on stage to talk about Siemens OP Center intelligence efforts.

Now, all of this, all of us in this room have a shared goal to make fact based decisions. And to do that, we need to put data to work. When we put data to work, we're able to identify new opportunities, improve efficiencies, respond faster to issues that need our attention. And we do that also in the context of organizations uh consisting of humans and human activity and how that impacts the world and the resources that we have. And so being able to measure that and test hypotheses uh requires storing data, managing data and that's how we put data to work.

Now, a data driven organization in the cloud era uh are organizations that seek the truth by treating data like an asset. It's no longer the property of individual departments. It's really a change in mindset to bring data to more people and to bring analytics to more people and to do that, they set up systems to collect and store and organize data and make it available in secure ways and through applications that can consume it. And they're also looking for the next set of technologies like machine learning that will unlock new value from data uh to improve operational efficiencies, develop new products and revenue streams and build customer, build better customer experiences now.

Yet, despite all the optimism and promise of value that data can bring. Research shows that organizations are still struggling to leverage the power of data across their organizations. Uh so if you look at uh many of these measurements, it's anywhere around 20% of the organization that is actually benefiting from business intelligence. And there's lots of reasons for why this is the case. But I'd like to talk about one which is this notion of fragmented BI and how we got here.

So organizations are made of users with different analytical needs and constraints. You have executives, they need quick numbers on the go, you have functional leads, they need threshold alerts uh to take necessary action. You have analysts embedded in departments that do you know that build dashboards and do their own analysis. And then you have um application users who want analytics embedded uh in their applications. They don't necessarily want to leave the application that they're in, but they want to do uh analytics in those applications.

Now, the diversity of these BI use cases and how solutions were designed, they were designed to address these sort of singular needs that you see in the middle of the screen here. So um it's not uncommon to find organizations that have one or maybe even more than one system to run paginated reporting solutions. Uh you may have acquired a company and therefore have another solution in house. Uh but oftentimes you have one or even more of these systems. And then through changes in technology, we saw, you know, we, we went from big CRT monitors to screens that we can put up in operation rooms. Um and we have tablets and phones. And uh organizations started to build up self-service BI capabilities. Um and so now they have another system designed to support those use cases. And even more. Uh presently, we find that organizations are also looking to deploy natural language query systems to their users and to do that, they've adopted yet another solution. And then finally to embed analytics in their applications, uh they have adopted yet additional systems to do those things

Now to add to this complexity. Keep in mind that each of these systems has its own connectivity layer for data, its own data modeling layer, its own calculation layer, its own visualization layer, uh and even their own scheduling mechanisms. So we're doing a lot of the work more than once and creating a lot of inefficiencies in doing so.

So Quicksight is our solution to fragmented BI. Quicksight enables BI for everyone in the organization uh that needs it in the format that makes sense for them. Uh Amazon Quicksight, we like to say offers BI at any scale or what we would like to call hyperscale.

Quicksight is a unified BI service that delivers modern dashboards for interactive analytics, paginated reports for agreed upon formats of data, embedded analytics for consuming insights directly in applications, and we have natural language query capability with Q that lets users just ask questions um using the language that they use every day.

Now, organizations can migrate away from maintaining multiple legacy BI systems for different needs and eliminate complexity. With Quicksight, Quicksight is built with builders in mind. So whether you are a BI analyst or a developer, we enabled you to get your analysis done and delivered anywhere quickly and efficiently. And there's no software or services or servers to install or manage. Uh customers can just simply lower your costs by eliminating investments in the necessary resources. And then with usage based pricing, uh you reduce your spend by simply paying for what you use. There's no long term contracts or hidden costs for extra servers.

So instead of struggling to work with disjoint legacy systems, BI authors and developers can use one unified modern solution to author the entire experience from the data connections to data preparation, to analysis, to dashboards and insights delivered in multiple forms, whether they're dashboards, paget reports, natural language solutions and even embedded solutions. And you can do all of that with Quicksight.

So the key benefits of Quicksight is that it's one BI service you need to access all of your use cases. It lets you consume insights and the formats that you want again from paget reports to interactive dashboards, to embedded analytics, you can do it all from one system.

Quicksight enables you to embed easily and seamlessly. We have one click embedding and you can easily embed analytics into portals and websites. Developers can further blend in the look and feel to provide a seamless experience to your users uh with the Quicksight SDK and to make it easier and therefore possible for more people to get value from data users can ask questions in natural language with Quicksight Q and Quicksight Q will translate their question and provide a correct response with a supporting visual and even supporting narratives with machine learning powered Q and now with Auto Q that reduces the overall time to solution uh deployment and populating these uh Q solutions becomes even easier.

Now, Quicksight is built on a highly purpose built serverless architecture that scales efficiently and instantly to hundreds and even thousands of users, no software or servers to install or manage and customers can lower costs by eliminating investments in unnecessary resources.

Quicksight supports enterprise wide secure sharing with governance and security. We have row level security column level security and you can define it in one place for all of these different scenarios.

And of course, you get more with less. By reducing the cost through pay per use. You can enable more analytics capabilities throughout the enterprise. You can save money in certain areas and spend it in more analytics use cases.

And to date, we have over 100,000 customers who are using Quicksight. And when we look at the actual outputs that you see in a traditional BI deployment, we see really four main categories.

The first one is this notion of building and deploying modern interactive dashboards that are designed for the screen. And by screen, we mean that could be a tablet, it could be your web browser, uh it could be your phone.

The second class of use cases revolves around natural language query where customers want to uh deploy, deploy analytics to even more users uh in their organization. And to do that, they're, they're, they're turning to technologies like natural language query uh to make that possible.

The third use case that we see is around embedded analytics. Uh so customers want to be able to embed analytics in both their internal facing and external facing applications.

And then finally, there's the use case of paginated reporting that we're going to dive into a little bit more today.

So the Quicksight team is constantly working backwards from the needs of our customers. And we're relentlessly focused on delivering functionality to solve your use cases. And to date, we've shipped over 80 new features into Quicksight this year in 2022. And it even includes really large pieces of functionality within that 80 per uh the those 80 new features. You know, the idea of paginated reporting is just one of the items in this list. Uh we've also introduced uh 1 billion row SPICE data sets as well as programmatic dashboards. But for this year, we're at about, we're a little over 80 new features for 2022 and we still have a little bit more to come.

So first, let's walk through this use case of building and deploying rich interactive dashboards. These dashboards are designed for the screen. They're designed for blazingly fast navigation, they're designed to be interacted with. And the data from these dashboards can be queried directly from the data sources where they reside using direct query or you can choose to ingest the data into Quicksight’s super fast parallel in memory engine called SPICE that again now supports up to a billion rows.

You can publish these dashboards with a single click and you can even schedule these dashboards to be delivered on a daily weekly or monthly basis. And in just a few moments, Rahul is gonna demonstrate how we're building out additional scheduling and snapshot functionality, not just for dashboards but also for page reports.

Now, I wanna talk through a couple of new features that we've added to the dashboard capabilities of Quicksight. The first is around the area of visualization. Uh this year we introduced a lot of functionality around one of the most important types of visualization which are line charts which are great at showing trends over time.

So we've added the ability for you to, to give you control out over how to display missing data points. We've added new customizations for markers. Uh so you can choose the shape, the size and the color of these markers. And then we've introduced new line style options. They're just not about visual representation, they actually serve very specific functions.

Um step line charts, for example, are a great way to show data that changes at a uh um you know, at a specific moment in time, right? So if you imagine something like the minimum wage in the US, it doesn't gradually go up over time, it changes immediately. And so you want to be able to draw that change with a perpendicular uh type of line uh visualization.

We've also increased the data limits for charts for line charts. So they now support 10,000 points.

Now, I also want to highlight the the new smooth line chart that we have that really improves readability for cases where you have overlapping series in your line charts. You can imagine if you're looking at a traditional line chart with very jagged edges, it's often difficult to know the difference between are these lines overlapping or is this the same line that's just changing direction? And smooth lines remove that ambiguity and they make it much easier to follow the lines as they're being um as they're presented on the screen.

And then lastly, we've added a new experience that lets that makes it easier to assign colors to series for charts.

We're also introducing small multiples, uh small multiples as you see here on the screen are great for comparative analysis and make it easy to digest and, and see patterns uh in your data, they're often preferred over stacked type chart variants uh because they really break things out into individual panels and our eyes just naturally begin to start to make comparisons when we look at, when we look at these charts and these are supported online bar and pie charts today.

Now shifting to natural language, Quicksight Q is our natural language query query capability that allows anybody to ask business questions um simply using the, the, the, the natural language that they are used to using in uh in the traditional conversation.

So here you can see what is the weak uh over week revenue difference. And behind the scenes, we're using machine learning models to interpret the user question and intent and we return back a visualization that provides an answer for, for the user. And again, this is, you know, sometimes the dashboard uh doesn't answer the question um you know that, that you have and so you're able to go to, to Quicksight Q uh and, and, and ask that question and get a response.

Uh but it also just generally makes analytics more accessible to people who, you know, simply, you know, the simplest user experience, you can provide is really just the question that you have in your head. Um and if we can build that as a user experience into our software, it greatly reduces um the the barrier to entry for folks.

Now, with respect to Q, uh we introduced new question types for Q here recently. So now you can ask forecasting questions on up to three measures uh just directly right there. Uh as part of your question, you can also ask why questions that will help identify key drivers uh in your data.

And this week, we also announced pin boards that allow you to save and share cue responses. So once you ask a question and you get a response, you can pin the visualization to a pin board and you can return to that at any time and share it with other people.

And lastly, we're taking the process of preparing data for natural language query and dramatically reducing the amount of time it takes to perform this function. And so we're using machine learning to go in and classify dimensions and measures, create name labels and then enrich topics with synonyms, semantic types and default dates. And this really reduces the overall time that's required to get a natural language query solution up and running.

Now, all of these capabilities are available in embedded analytics in your app for your applications. Uh you can embed dashboards natural language query experiences and even dashboard authoring into your applications, it's super fast to get started. There's no custom development or domain expertise needed. And again, there's no service for you to manage simple maintenance. Uh it scales to whatever your needs are.

And just this week, we announced new expanded APIs that give you programmatic creation and management of your dashboards. So this is incredibly uh this is super critical in terms of being able to integrate your dashboard and dataset assets and all of your Quicksight assets into your dev ops situations where you have a code that's being stored in repositories and you want to back those, you want to back up these assets.

Um you can now do that with these new APIs and the data that they return. Uh you can also use these APIs to accelerate your own migration from older BI systems where you can interpret files and create Quicksight assets. It really reduces the time for you to accelerate those uh for you to do those migrations.

And of course across the entire portfolio of capabilities, we're augmenting data with machine learning. So we have auto narrative capabilities that will summarize business metrics in plain language. Uh we provide machine learning predictions directly in the product. Uh and we also enable you to connect to Amazon SageMaker models"

uh so you can call those models that are deployed elsewhere and bring those results into quicksight. again, we're using machine learning with respect to our natural language query capability and where we have machine learning available to do anomaly detection. so you can discover unexpected trends and outliers in your data. and we also use machine learning in our forecasting capability uh that's built directly in the product and we saw some of that with q as well.

now, i'm really excited today to talk about an entirely new use case that we're bringing to quicksight and that's paginated reporting. now, paginated reporting uh is a critical uh capability. uh that's been in the, it's been, it's been in the industry for decades and it's exactly what it sounds like. it's paget page based reporting. and so i'm going to talk a little bit about what multi paid or what page reporting is and then talk about the benefits.

so the first is that page nated reports, build multi page reports, you can generate pdf reports that render all of the data, uh all the rows of a pivot table when it's exported out. and you can see an example of this here on the screen and you can see how it's different from a rendition. that would be, for example, in a dashboard in a dashboard. if you have thousands of rows in a table, you might simply scroll those rows, you uh those rows, you might uh use a um you know, a u i gesture to be able to click next, next, next. but when you go to paper and when you go to pdf, you want all of the data. and so we have to give you control over how that data is laid out. and so we're giving authors the ways to define explicit page breaks and control the flow of the output headers and footers are also a really important aspect to reporting. uh you need to be able to add parameters for things like page numbers and total number of pages in the report. um and you want to be able to add logos and visuals and other text to the header and footer areas. and so we're giving the a, we we have the ability to for you to create headers and footers with page reporting.

and then quicksight page a reporting also includes scheduled delivery so that you can schedule pdf documents that span multiple pages and you can also export cs v and we're adding new snapshot history capabilities so that you can pull up these snapshots on demand. so we'll store all of the report snapshots and you can pull these up on demand.

now, the benefits of pageant of reporting in the cloud are that now we have this additional use case such that you can build all of your insights independent of your preferred consumption model, whether it's a dashboard, whether it's report, whether it's through natural language on a single source of truth and governed data sets. so that calculation that you do for profitability, you can define it once and you can use it in reports, you can use it in dashboards and you can use it in natural language and you can use it in embedded solutions as well. it's not multiple data stacks, it's not multiple calculations, it's one set of data and you can consume it how you it's best for your users.

it's also with page and reporting. the authoring experience is built directly into the quick site analysis, authoring experience. and so we're introducing a new sheet type. there's nothing new to learn. it's the same charts, the same calculations, we're just adding new options and new functionality in addition to what was already present. and again, you pay for what you use and there's no software updates to manage.

so with that, i'd like to invite rahul up to the stage. he's gonna do a deep dive demo on pageant and reporting. so come on up sean and uh there's the clicker. all right.

thank you everybody. um so before we get started with the demo, i wanted to go maybe set the stage um for what this demo is gonna be about. so imagine if you will, each one of you are quick site authors you're set out to build. uh and your job is with coffee emblem company. ok? you saw a couple of examples already. so you are responsible for building such beautiful dashboards. that's what you do. day in day out. the coffee emblem company has just decided that they want to sunset their legacy application, the legacy b i tool, and one of the critical reports that your job is today is to rebuild one of those reports. within quicksight, this report is gonna be distributed to sales reps all over the us on a monthly basis. and it will contain information such as what are the top selling products. how are you doing in terms of market size and detailed information about what products are selling well in what states? so with that, let's take a look of how this works.

so what you're seeing on screen is the quick site u i, this is a web u i web-based. um there's nothing to install, nothing to do. and what you will notice is that i've already taken the liberty of favoring the two objects that we're gonna be working with today. the first one being analysis. so this is how we build dashboards within the quicks site tool. when i click into it. what you'll notice as it comes up is that this is a dashboard that we are utilizing today. it highlights a couple of the visual types. we have like geospatial, we have a donut chart, we have a waterfall chart and we also have the insights here. it identifies automatically what are the top three products as well as how does our profitability compare to last month? as an example, we also have uh a bar chart as well as the tabular data off the information itself. and that kind of sean was mentioning when it's in a dashboard, it's ok. in this kind of form, you're able to kind of scroll through and navigate and identify what are the rows that you wanna kind of drill in on. but let's get back to the manner of hand, right? we wanna build a report, so we wanna use the same exact data set, but we want to just take the dashboard as this and this kind of convert it into a report. and that's kind of what this functionality enables with this one click. what happens is it converts all moves over all the visualizations we had from the dashboard into a report allowing me to now delete and remove visuals that may not be applicable for this report. use case that i have, i'm able to use quick sites, drag and drop capability to place visuals where they're most applicable. i'm able to resize them so that they're visual and readable. and we're introducing this concept of a section, what a section is it, it helps enable a sequential order in which you want to place these visuals. so i've just created a new section dropping in this table into this section. so now the system knows, all right, let's just render the visual one to the bar chart and the waterfall before let's moving on to the table below it. as you note in the bottom right of that table it says it prints across multiple pages because it's not just gonna render what is shown on a section level, you're able to define explicit page breaks. so after these two visuals have been rendered, there needs to be a page break. and then the next set of things start, you're able to obviously remove that and add sections in between the sections themselves move as a whole, right? so now you're able to move a collection of visuals between before each other and remove. so a header is nothing but a special section for us, right? it just repeats on every single page. so let's move along in our reporting journey and we're gonna add our coffee emblem logo. so we're gonna use the custom visual option and i'm gonna go in and add the coffee emblem logo and resize it so that it repeats on every single page of my report. we're also introducing a net new visual type here of adding a text box. this allows just free form entry of text without any other uh frills or any other details you're able to define a rich text experience. so that way you're able to move this report title. so now let's gonna move down to the footer, right? so what do you typically associate with report footer page number and when the report was last generated? so again, we're gonna lean on our text box where we have provided an easy discoverable, quick way to get to those both page numbers as well as printings. so here as i click through, you can see that there's predefined set of formats there, you're able to just with a single click, be able to define that this is sales data after all. so it has to be confidential. so let's go ahead and mark this report as being confidential. so every time it gets printed, the footer will display it as such and then let's not forget about a page number. so again, going back to our text box, we're gonna add the page number to this photo. so at this point, i will take some liberty and kind of do the food network style and say, ok, let's take a look at one that i prebuilt.

um but before i do so, i just want to kind of call out one last capability, which is that the settings that you see on the page or the canvas is completely customizable even after the fact, right? so as you're building it, you may realize, you know, i'm running out of space or this table is too large. i need to switch from portrait to landscape or vice versa. but let's go ahead and take a look at what i prebuilt without too much further time and a little bit of formatting elements that are already prebuilt within quakes site. you can see that a couple of color changes and it looks pretty beautiful even if i say so myself here at this point. let's imagine you didn't have a dashboard you wanted to start from every time you click a new sheet. this option presents itself, right? so here the author is presented with a choice. do you wanna build an interactive sheet, which is a dashboard for us or do you wanna build this pageant or report? and you the power here is, are you able to build them side by side in the same analysis experience using, using the same exact data set? right? so here this is a blank sheet. this is world is your oyster, whatever you want to build, uh you're able to build.

so now let's get back to the report that we just built and let's try to complete our job by publishing that to a dashboard. so again, here at this point, we've introduced a new functionality whereby at the time of publishing you are given. um so you're either able to publish as a new sheet or an existing sheet. but then above and beyond that, you're able to specifically select which sheets you want to publish. so here you could still have work in progress sheets and things that you don't want the world to know yet. so you're able to selectively choose and select which ones need to be published. so let's go to the dashboard as it's published and we selected two sheets if you recall, one was a dashboard sheet and one was a report sheet, the dashboard sheet as you imagine and as you probably would expect aren't interactive. you're able to zoom in zoom out of this data, you're able to drill in, you're still able to navigate through the table via the scroll actions. but the cool thing is now you're able to go to the report sheet and now you're able to satisfy that use case that we started off with. i'm able to now send on a monthly basis, a snapshot version of the report. in this case, a sales report. and what you're seeing on screen is that snapshot and you're seeing them side by side, right? so imagine i'm a user, i wanna see what did my data look like as of december? and that's what this view is. i'm also able to via this drop down, go back to a certain point in time. what did my data look like as of november this month? and that is an immutable version of that truth. you can see that the best selling drinks are different, but you're able to kind of zoom in zoom out, go to specific page numbers, navigate through the snapshot that has been created and you can see that it's multiple pages. it honors all those different things that we did page numbers inclusive, but that doesn't stop there, right? so let's move to the recent snapshots area where you're able to not only see the pdf and download the pdf that we just generated, but you're also able to download the cs v equivalent of that table that was there. so now you're able to satisfy your users by not only providing a highly formatted output but also the raw data that they can then consume in an excel or downstream systems.

so here we've created the ability to um schedule, create multiple schedules. and if i drill into one of those schedules, you can see how that was configured by providing a schedule, name and description. the author is able to select what content they want to be able to distribute. pdf allows you to select which sheet you want to choose dashboard sheet or report sheet. but then the cs v, it allows you to select specific visuals from that entire dashboard tables, pivot tables to start with, you select the cadence in which it should be distributed daily, weekly, monthly and you're also then selecting whom it should be sent out to. so providing a subject line recipients, and then you're able to specify if you want to include a preview of the report itself in the email body. for pdf, you're able to send it as an attachment or if the data is sensitive. let's imagine it's ppi i or hipaa compliance. you can then send that as a download link. so that way you're not physically having an attachment floating around. so in this particular case, let's go ahead and save, let's have, wait a second or two and imagine that it's the end of the month. and voila, you've got an email as a recipient. so this is our coffee emblem, month to date sales. you have the attachment of the pdf, you have the link to the dashboard that generated that pdf as well as a one page quick snapshot of what that sales number looks like, right? so you're able to quickly glance on your mobile device wherever it may be. and then if you want to dive right into it, you can, if you click on this view, recent snapshots, what it does is a quick n a, it goes opens up the dashboard and it also opens up the recent snapshots. so you're not now constrained to just looking at what was just generated, but you can go back historically and download and view what was generated last month or the year prior.

so let's look at a couple of other examples. um so this is an example of a financial statement that we generated. again, if you think about your own pers personal use cases, you can easily imagine finance users asking for specific predefined look and feel you want the ability to add headers and footers, discrete page breaks. um and imagining of a catalog if you may of um a table with visuals that accompany it, so you can see the stark contrast to what a dashboard is. in this case, we imagined if we had a quick site university right. um then you have students with grades and how would they do? um so you're able to not only design tabler data, but you're also able to leverage existing visual and visual types in the same look and feel here. our designers went a little bit uh outside the box, but you're able to kind of again imagine the ability and the endless possibilities that this pageant reports to us. and then last but not least, let's end where we started off with, with our monthly sales report highly formatted. it honors role level column level security and it look and it can be distributed out to the masses.

so at this point, i'm sure you're all wondering how do we get started and what do we do next? so, um the uh pageant of reports is an add on to the enterprise and enterprise plus q editions of quicksight

You can ask your account admins to go to the managed subscription page and be able to add on QuickSight as an add on. You are able to select between a monthly or an annual option for a commitment. We've intentionally kept the pricing simple. So there's no additional cost to authors or readers and it's again, more from a usage based model that we're going.

So, on that note, I would like to turn it over to Max from Siemens to help talk through how Siemens uses this.

Yeah. Thank you. Thank you. Thank you very much. My name is Massimiliano Ponticelli. Yeah. So I'm really happy to be here to talk about my product Siemens Digital Industries Software and how we can solve the issue of our manufacturing customers with our AI analytics together with the AWS services and of course, the AWS QuickSight.

So starting from the beginning, we are part of the Siemens Digital Industries. Siemens Digital Industries is the market leader in the industrial software. We are focused on complex realizations to leverage the IoT and to move to digital transformation to the companies from all the value chain starting from the design phase to the production.

So let's consider that the data, it's really the important part of this journey because the data coming from the production and the product can be used, analyzed, can be used in the engineering part to improve the product and to then improve the whole value chain in a very effective way. So data is really the center of our organization.

What can the customer do with our comprehensive digital twin? They can predict product and performance of the production. They can increase, of course, the throughput, the speed of the line. They can reduce the time to market. They can improve the quality as well as they can improve also the design, the design phase and for example, reduce the physical prototypes using the digital part of our digital twin.

Siemens Digital Industries Software and Siemens Digital Industries portfolio elements are part of the manufacturing. So we are mainly dedicated to the production part to improve the production part.

What our customers typically are asking, mainly today, talking about pandemic, talking about crisis of energy, they are affected from fast and low cost production. So they need to improve the efficiency. They need to face high demand variability in the customer requests. Of course, they need to reduce the time to delivery and they always need to keep a high level quality standard.

Of course, there are other problems like the supply chain, you know, the supply chain in this time is very different from time to market to time to market. There is high consumer demand, there are high consumer costs, there are low and slow deliveries and all this kind of stuff can be solved by digitalization.

What does it mean? It means that of course together with our portfolio, we need to bring together flexible technology and this of course is cloud, IoT, local development and of course analytics. So, analytics is really the core of the digitalization journey for our customers.

But we know and that to be in this digitalization, the customers need to add intelligence, equipment, a lot of sensors, a lot of data. And we know in this room that a lot of data does not mean to be able to analyze the data. So you need a level of intelligence at the top of this data because having a lot of data is really important, but you need something more.

That is exactly what we built with the Siemens Digital Industries Software, a layer of intelligence, sorry, for intelligence that is able to transform the big data coming from the IoT and also the data coming from the process and the production environment to a sort of smart data set with a holistic view across the different processes in the manufacturing.

That is of course having this layer of intelligence and data. That is the first step. But how we can deliver this data in an effective way to our customers, our users?

So in manufacturing, we have of course different requirements. So the users want to have an easy access to the data. So they, they don't need to, they don't want to be an expert of making SQL queries or making very complex stuff. They need to have a tool easy to use.

We want to embed this tool in our Siemens portfolio. Of course, our customers want to have a single point of access and see all the data altogether.

We have a very strong need of data segregation or level security. We have an enterprise organization - users are different from location, from department. There are enterprise users. So our customers want to segregate the data and give the data to the right person, to the right decision maker.

As we started at the beginning, our users want to have flexible dashboards and KPI analysis, drive and data. And that is true - we started from that point of view. But now, our manufacturing customers are always asking for print reporting. They always ask and they have a couple of examples in the next slides. They always are asking to have a way to paginate, to print all the data, all the parameters, to export in PDF, to save for a long time.

And as well, of course, the scalability for the manufacturer is super important because we have small medium businesses where of course they want to be conservative in terms of cost. We have big enterprises that will start small and grow in time based on the business.

How we have built our manufacturing data hub on AWS is exactly with this high level architecture. So we have different kinds of data. We have real time data coming from IoT. We have more structured data coming from enterprise systems or production data. And of course, we are using different services.

So you can see we are using for real time, we are using Kinesis, DynamoDB as short term data store. We are using RDS and of course, QuickSight being natively integrated with all these services is really the point to connect everything for our users.

And even more today having the reporting, we can keep our architecture as it is today while adding a new important feature for our workflow.

Now, this is really a focus point for us starting from this architecture. I want to expose a couple of examples of our customers. One is using right now the dashboards and moving really quickly in the next days also to the reporting.

Vorwerk is a German company leading in the household appliances. It started considering having a solution putting together PLM, ERP and shop floor data. They started with a solution with more than 20 different data silos coming from different systems. And so they asked us to have a single point of truth, having all the data in a single point and easy to access for every user.

And they are using QuickSight right now on top of our architecture for different kinds of users - use cases. They have QuickSight in the factory, in the line, in the big screens where the line supervisors and operators can see how the data are moving in near time and taking tactical decisions. And they have QuickSight also in the offices where there are power users that can take the decisions in a strategic way.

The other example is Absolut. I think you know Absolut vodka. It's quite a famous brand. And also being a famous brand with a very wide market, they need to keep under control the efficiency and quality of the system.

So also in that case, the requirement was to have a system that was able to have better traceability of the data, having better order management and also having a system that can react quickly on any losses in production or quality.

Also mainly in this case, having a very strong requirement of traceability, the reporting is even more important because they can trace all the data. They can have reports, they can save in PDFs, they can go to QuickSight every time and see what happened in the past.

And of course, we have our plants with our products using QuickSight in our facilities.

A couple of examples of common reporting in manufacturing:

1. The production efficiency report - in this report, typically the customers are tracing all the activities in the equipment - what is the changeover, what kind of faults, what were the parameters or the energy consumption of a certain production? And they are using this reporting in what they call daily meetings or shift meetings. So they are printing the reports, they are using them daily or before a shift to take technical decisions for the next period.

2. Another super important one - I have plenty of examples - but another super important one is the production batch report. This is really important because it is tracing all the activities in production - each step of production, each rework, material movement, any loss of data, any comments from the operators.

And this is in some industries like medical devices really important because there are regulations like FDA for quality that require you to save and keep the data for a long time. This is really used to deliver the final product in real life. So having this is a focus for some industries.

To summarize and wrap up the main benefits from what I discussed, of course, with my team and also with two of our architects here:

From our point of view as a product, we recognize big value in a PaaS product like QuickSight in terms of cost, scalability, deployment, updates - because we don't need to do anything, AWS is doing it for us. There are also a lot of savings for us from the architecture point of view.

Of course, having new features like reporting that we saw today is super important - we get those features for free. We don't need to reorganize our architecture every time to release something new to our customers - that is very important.

Also, from the time to market perspective, having a PaaS, having a rich set of SDKs and embedded capabilities, we can deliver quickly out of the box functionality for different industries to our customers.

From the customer point of view, there are mainly two points:

1. Having a single product with dashboards, machine learning and reporting is a big value - our customers and users don't want to learn different tools. They are operators, they want a simple solution to learn, they don't want two tools to make reports and dashboards. QuickSight gives them that.

2. Cost and scalability - being able to start small and scale as the business changes is very important, especially today with the pandemic. QuickSight gives them that flexibility.

That is why we started with AWS and QuickSight, and why we continue the journey. We hope to have more and more customers together in the future.

Thank you. I definitely want to thank Max and his team for partnering with us on QuickSight. It's been great to build out reporting capabilities together and see their story - it's wonderful.

I want to quickly call out there is a session not listed here at 4pm over at MGM on "What's New in QuickSight" - a full tour of capabilities.

Finally, I encourage everyone to join the QuickSight community - QR code there. Product team is there, great content and demos. It's a great place to learn more.

Thank you. Please complete the session survey and enjoy the rest of re:Invent!