In this episode, we’re talking about the IoT (Internet of Things) space and the related data that drives the industry with two industry leaders from Crate.io who specialize in IoT data management: Christian Lutz, founder, and president, and Eva Schönleitner, CEO. Data in the IoT space is unique in many ways, most notably in shape and scale.
In this episode, we’re talking about the IoT (Internet of Things) space and the related data that drives the industry with two industry leaders from Crate.io who specialize in IoT data management: Christian Lutz, founder, and president, and Eva Schönleitner, CEO. Data in the IoT space is unique in many ways, most notably in shape and scale.
Also discussed are some of the many announcements made by Microsoft at the annual Ignite conference, which was held recently. As always, there was a lot of big news that came out of Ignite. Microsoft’s Diego Tamburini, Principal Industry Lead for Cloud for Manufacturing gives the scoop.
Episode Links:
Episode Transcript
Crate.io
.NET Core on GitHub
Percept
Azure Percept: Edge intelligence from silicon to service
Guests:
Christian Lutz is the founder and president of Crate.io. He has over 20 years of experience as a startup entrepreneur with strong experience in enterprise software.
Follow him on LinkedIn.
Eva Schönleitner is CEO of Crate.io. She has extensive worldwide experience in the technology and industrial segments and has a track record of developing high-performing organizations.
Follow her on LinkedIn and Twitter.
Diego Tamburini is Principal Industry Lead for Cloud for Manufacturing at Microsoft. He helps build the best ecosystem for software partners developing manufacturing solutions on Azure and helps them be successful.
Follow him on LinkedIn and Twitter.
Host:
David Starr is a Principal Azure Solutions Architect in the Marketplace Onboarding, Enablement, and Growth team at Microsoft.
Follow him on LinkedIn and Twitter.
DAVID: Welcome to the Azure for Executives podcast, the show for technology leaders. This podcast covers trends and technologies in industries and how Microsoft Azure is enabling them. Here, you'll hear from thought leaders in various industries and technologies on topics important to you. You'll also learn how to partner with Microsoft to enable your organization and your customers with Microsoft Azure.
Hello, listeners. Today, we're talking about the IoT (Internet of Things) space and the related data that drives the industry. So data in the IoT space happens to be unique in several ways, which we'll discuss throughout the show. Now, also up for discussion are some of the many announcements made by Microsoft at the annual Ignite conference, which was held recently. As always, there was a lot of big news coming out of Ignite, so we'll visit some of that today. And joining us are two industry experts from Crate.io who specialize in IoT data management. Firstly, let's welcome Christian Lutz, who is founder and president at Crate.io. He has over 20 years of experience as a startup entrepreneur with strong experience in enterprise software. Welcome to the show, Christian.
CHRISTIAN: Thank you for having me. Welcome.
DAVID: And Eva Schönleitner, the CEO of Crate.io She has extensive worldwide experience in technology and industry segments and has a track record of developing high-performance organizations. Now, I’d be remiss if I didn't also mention, Eva, that you're an ex-Microsoft alum. So, welcome to the show.
EVA: Thank you so much. I really appreciate it.
DAVID: And finally, from our team on our side, we have Diego Tamburini, the Principal Industry Lead for Cloud for Manufacturing. And Diego helps build the best ecosystem for software partners developing manufacturing solutions on Azure and helps them be successful. Diego, you've been on the show a couple of times. Welcome back.
DIEGO: Thank you, David. It's a pleasure to be back on your podcast.
DAVID: Let's just kick off maybe with Eva. And I wonder if you could introduce us to Crate.io. Tell us a little bit about the organization.
EVA: Sure, absolutely. What problems do we really solve? When you think about it, we're talking about digitalization initiatives here, especially in the area of IoT. And over the last several years, many companies have been emerging and really gaining a foothold. And today, this is really a large industry. In fact, the IDC just had a new study that says this year, AI revenues, AI machine learning revenues are going to be $327 billion worldwide with annual growth of over 17% per year to reach over 500 billion just in 2024. So this is a huge market that we're talking about, and we are part of this market. When you think about it, as digitalization happens with companies, a new group of technologies have been coming around, and we'll hear more about the Ignite and Azure announcements that you already mentioned. But also in terms of specialized best of breed databases, and that's exactly what creates iOS and some other tools and connectivity options, think 5G and so on.
So we are one of these technologies that’s specifically made for these new digitalization initiatives. And we're a database technology that is meant to be for these machine data use cases that you see when you have these digitalization initiatives in the IoT and in the industrial segments. So it extends not only to the typical database use cases. But when you're talking about machine data, it comes from a machine, but also non-machine data that are structured with unstructured data, unstructured means think pictures, think central data that are coming at a large scale at us. When you want to combine these data and then do analysis, particularly also in real-time, not just historic, but real-time, which means then in the category of time-series, that is when you need a specific database, and CrateDB is one of these databases and a top-notch one at that.
In terms of Microsoft alignment, we actually have a SQL front end. So in terms of usability, it's very familiar with all the Microsoft users, very easy to do. And we have an open-source back end that then allows this massive scalability and also the combination of the structured and unstructured data, and Christian, we'll talk more about this. We are also, of course, a very deep Microsoft partner, not just from a technology perspective from Azure but also from a go-to-market and marketplace perspective, just as a quick outline of where we are in the ecosystem.
DAVID: That's great. Thank you for that round trip. It sounds like you guys have a lot of industry expertise in IoT and the data surrounding it. So with that, maybe we can jump right into it with Christian. So, how is managing data for IoT different than we might see in traditional data management scenarios?
CHRISTIAN: Eva pointed out it's a mixture of structured and unstructured data, and the core differentiator is A, the scale of data and the shape of data. So the scale is obvious when you start connecting machines and your lines, then you will quickly grow from gigabytes into single terabytes. You may end up in hundreds of terabytes or even petabytes. With the shape, it's more complex. That's really the high variety of data that you need to manage. So, for example, you will have from an ERP system coming strictly relational data, but you also will have complex sensor data, time-series data, maybe logfiles, maybe even geospatial data, and all of that together in one table ideally. That means when you combine that, and you have the scaling requirement, for example, you go from one line to all the lines in a plant. And then one of our customers, for example, has 180 plants, then that's what we mean with scale. And so traditionally, multiple data stores are used to solve such a problem. And we think the simplicity of using obviously one data store is really the interesting characteristics that we can provide here.
DAVID: Oh, okay. So that's fascinating. You are proposing that all devices can report back to a single database instance as opposed to storing in separate databases and aggregating on the back end. Is that right?
CHRISTIAN: Yeah, correct. And the advantage of that is that you have less complexity, and usually, you have to then create merged tables and merged use between different silos, and that makes it a lot easier and also much more simpler to operate.
DAVID: Perfect. Okay. Diego, you've seen a lot of IoT installations in your life, I'm sure, and a lot of big ones too. So I'm wondering if you could talk with us about some common patterns for data storage that you see in IoT computing.
DIEGO: When it comes to IoT solutions, it is common to refer to the data taking three different paths. Obviously, this is a generalization, but it helps sometimes architect IoT solutions. And these paths are depending on the frequency and speed that the data has created and consumed and how much transformation it undergoes before being consumed. So the three parts are a hot path, a warm path, and a cold path. The names imply that a hot path refers to the data that is immediately used. It's more characteristic of the streaming type of analytics and data storage, which Crate is so good at as well. And the data is constantly analyzed as the data is coming, so streaming data. And then, on the other end of the spectrum, which is the cold path, is where the data is structured. So it's used by reporting; it’s put into data warehouses and things like that. So it undergoes a lot of transformation and structuring before being used. But like Christian just mentioned, a database like Crate.io is very flexible, and it can handle the storage of the data for those three data paths.
DAVID: Thanks, Diego. That’s really interesting. And I wonder, to go along with that, if you could take us on a brief tour through some of the more significant Azure IoT services. What are some of those services? And maybe you could briefly describe them for us.
DIEGO: Sure. The heart of our IoT solutions is the Azure IoT Hub, that's our IoT gateway if you will. It's a managed service that is hosted in the cloud and acts as the central message hub for communication between the IoT application and the devices. And this communication is bidirectional, from the devices to the cloud and from the cloud to the devices. Then there is the Azure IoT Edge which is the counterpart, if you will, on the devices. So basically, it allows you to package business logic into standard containers and deploy these containers on the device to do some local processing. And the pattern is that if you need something done and processed very quickly, you do it on the edge, and then you send summary data or particularly interesting events to the cloud. Then there is the third one is Azure Stream Analytics. Azure Stream Analytics is a fully managed real-time analytics service that is designed to process streaming data, what I just referred to as hot data or data through the hot path. And fourth is the Azure Time Series Insights. So the Azure Time Series Insights is a full-fledged application to store, visualize, and query large amounts of time-series data. So again, IoT hub, which is the heart of our IoT solution offering IoT Edge, Azure Stream Analytics, and Azure Time Series Insights.
DAVID: Thanks for that, Diego. It's a great trip through those technologies and product offerings. I wonder to Crate.io, whichever guests would like to answer, how does CrateDB compliment Microsoft Azure's IoT services?
CHRISTIAN: So we're really fitting quite neatly and also natively into this environment with a direct connection to the IoT hub, for example. So you can simply ingest the data. There is a rich offering on Azure, especially on IoT. And we have focused on a certain niche that we see between the Azure SQL offering, TSI, and on the upper end, how we would call it Cosmos DB. And there is a white spot for a highly scalable database that really focuses on machine data use cases. That means huge volumes of, let's say, cheap data, not transactional data. Cheap data in the sense of billions of rows of sensor data, time-series data, and there to have an offering that can cost-effectively manage this load, offers a SQL interface and is cloud-native with a pay-as-you-go model as everybody's used on Azure. That's essentially why we think it's perfect for large-scale IoT use cases as a complementary solution to the Azure IoT offerings. And if you look to our production use cases, that's also how it's placed. It's placed in the middle of that stack, and through the SQL interface, it can super easy connect to the Postgres driver to any surrounding applications, e.g., Power BI or Stream Analytics and other offerings on Azure.
DAVID: Perfect. Okay. That makes a great picture for how you fit into the overall ecosystem. I wonder now if we could turn to maybe some stories about implementations with partners or customers. So, is there a story or two that you might be able to share about organizations using real-time data streaming and data visualization?
EVA: Absolutely. Let me give you an example with one of our customers in the IoT area, the company is called ALPLA, and it’s actually a multinational global bottling manufacturer. So basically, think of a manufacturer that produces plastic bottles. Think of a very famous drink company based in Atlanta and their plastic bottles, or also, if you go into the grocery aisles and look at our soap bottles and many other plastic bottles there, that's where those bottles come from. And so that company has 180 manufacturing plants all over the world, including the U.S. And what we're doing there is today we built an application that runs completely on Azure in the cloud. And what it does is it actually interfaces with the live production while these plastic bottles are being produced, captures numerous sensor points while in production, and then monitors the output of these machines. And so there's a capture, there is an analysis, and then there is an action on top of that. And the action is as the machines run, the company is able to put as little plastic as possible in there and still be within the guidelines. And why is this important? Because in this case, the plastic granules are the most expensive part of this bottle. And so if you have less plastic, your cost of goods goes down. Obviously, also if you can run it at a very high quality, then your quality and measures and metrics are much higher, and your bad items are much lower.
And so what we do there runs at 16 of the factories today. As the bottles get produced, you basically track automatically how the machines basically produce the bottles. And if anything runs away in the wrong direction, up or down, or a machine is running into troubles, there's an analytics engine that basically then immediately alerts the workers from the clouds down to the plant onto their smart device into their Bluetooth headphones, which are also earmuffs and basically tells those workers in the plants “Go over there to line number eight, third machine, and please be sure that there is a trouble. Please do make some adjustments.” And so, what this is, is all real-time massive amounts of data, and it saves not only on cost of goods sold in terms of plastic, ingredients but also obviously brought the quality of the product up and also the company was able to do a significant labor savings and deploy some of these workers to do some other tasks within these plants.
DAVID: Just to drill down on that just a little bit, that's a fascinating story, not the least of which is the amount of tech involved in making that round trip with that data happen and the alerting back into devices. That’s just fascinating. So it sounds like it's the case that this solution has actually benefited the company in being able to have greater uptime. You mentioned some of the employees being able to go work on other duties. So is that right that they're actually improving the manufacturing line?
EVA: It is right. It reduces the amount of plastic they need, so their margins go up. It brought the line time up. It reduced the bad output, basically increased quality as well, and the workers became more productive. It has a huge OEE, double digits, and that is why we're doing this. And he said it all runs in the cloud 100%. It all runs on Azure which is real-time. It's extremely impressive.
DAVID: Very nice. Diego, how about you? Is there a story you might be able to share about some interesting scenario with real-time data analysis?
DIEGO: Yeah. It's from a few years ago, but it's still one of the most interesting things I have seen of high-speed edge computing from a machine from an industrial manufacturer, a customer of ours, that is used to detect the presence of a highly carcinogenic mold called aflatoxin that is present on kernels of corn. And the scenario was that the corn gets fed from a truck into a hopper. And then, they feed a sensor into a chute. When it accelerates to 3.5 meters a second, you cannot see the individual corns with a naked eye. And then a camera on each side UV lights to eliminate the grains, and it’s looking for a particular fluorescence signature of the mold and with AI running on the edge on the machine itself. So when it attacks the mold signature, it shoots air via high-speed valves that should compress air and rejecting the individual corn kernel that is infected. I saw this machine operation at Hannover Messe about three years ago, and it was absolutely mesmerizing to see the machine shooting the individual infected kernels of corn. So to me, that's one of the most interesting examples I've seen of AI running on the edge and having to be extremely fast.
DAVID: [chuckles] You're right. That is very intriguing. I can only imagine the amount of tech that went into building that and the amount of effort. That's incredible.
[Commercial]
And now, let's take a moment to hear a very important message. The Azure Marketplace and AppSource Marketplace provide opportunities to sell your Office 365 or Azure-based solutions to customers in over 140 countries, giving you a new distribution channel for your products. Selling your products in the Marketplaces brings benefits in terms of partnerships and collaboration with Microsoft as well as go-to-market benefits to help you be successful. To learn more, visit the following URL, aka.ms/marketplace, and discover how to be successful selling through the commercial marketplaces while generating hundreds of new leads.
DAVID: So, turning the conversation just a little bit and steering it in a different direction, I'd like to talk with Crate about the open-source strategy that you guys have developed. And this is a big topic for a lot of organizations typically consuming open source. Some companies also will build an open source-based product and have enterprise features on top of that that they might charge for. But for you guys, as of CrateDB 4.5, you're releasing your source code as an open-source project, even the enterprise features. So many companies typically hold back on those value-add features and sell them as add-ons to a base open-source product. So, Christian, I wonder if you could tell us a little bit about what led to the decision to open-source the enterprise features of CrateDB.
CHRISTIAN: I'm happy to. Open source has been, since the foundation of Crate, a cornerstone in our strategy, and that's essentially how we built the whole product. It stands on the shoulders of some very large open source chains. There are many benefits to it, including our existence, really. So, for example, large enterprises can look into the code quality. It allows contributions of the community. It helps us to get adopters or interfaces more quickly than we otherwise would have maybe time or resources for. And we always have been built on that with an open core model and a closed enterprise feature set. And now adding the enterprise features also to the open-source codebase that essentially comes from us realizing and recognizing the fact that most of our paying customers really use our Azure offering and the management services we give them. And they don't really pay for these individual carved-out features. We see ourselves really as part of the modern data stack, and open source is really an important characteristic to be part of that offering.
DAVID: That's really interesting, the tack you've taken there. How does open sourcing your entire stack fit into your business model?
CHRISTIAN: As I mentioned, already most of our customers, it's really almost 90% are using our cloud-based offerings and pay us to actually run the service for them and most effectively help them to optimize cost and availability for the services. On-prem customers for self-deployment also want to have access to our SLAs and the support 24/7 and all of that. And we continue to grow with that, of course. But overall, we don't really expect a big impact on our business model because those people who don't want to pay, anyway, don't pay. Those who value the support and who want to run the service in the cloud as a pay-as-you-go service, for example, they value a lot what we do. And of course, overall, this whole thing simplifies our own internal development of code; there's only one codebase now. And the same applies to our partners and customers that built this technology into their offering and which makes it risk-free and also simple to distribute based on their Apache 2.0 license.
DAVID: Hmm. That's really interesting for businesses that might be looking at an open-source strategy. This is one that goes a bit beyond what we usually see. So in that vein, Diego, I wonder if you see many partners with open source products offering their solutions through Azure, particularly the Azure marketplace.
DIEGO: Absolutely. We see that all the time because of multiple reasons, one, because developers want to develop an open-source. They feel more confident, and that's where they learn to develop and also sometimes because they want to be platform agnostic. So Azure, as we say, loves open source. We've embraced open source because, like I said, it comprises most of the codebases out there. We support open sourcing programming languages, operating systems, databases, AI platforms; you name it. So most of the popular platforms out there we support and completely embrace on Azure. It's now part of our DNA if you will.
DAVID: That's a perfect way to describe that. Working at Microsoft, of course, we see that all the time steeply ingrained into our culture today. Let's move the conversation over to edge because we mentioned we would talk about that. So when we talk about IoT today, we have to include edge computing as a topic. And that typically means geographically distributed devices that capture and process data without sending that data back to the cloud and operating on it locally. So, maybe to Eva, how about storage on the edge? How does that work?
EVA: Thank you. You already captured it before. We don't even need to explain why the edge is important. But just to recap, we had on-prem for a long time. Now we're on the cloud, and pretty much everybody's using the cloud. But when you talk about the use case I just mentioned, it's a massive amount of data. And what I just described was only 16 plants out of 180 plants, and it was one customer. So when you look at the development and the latest study I've seen, is that the forecast is 99% of all data will be on the edge. My personal belief of why somebody is saying this is because the data on the edge is exploding exponentially, and that's why there is so much. And so, by default, you can't send it all to the cloud. We also see locations where the cloud connectivity, especially in remote areas or certain countries, let's say in Asia today; it's just not fast enough to really deal with these real-time use cases I just mentioned. And so you have to do some compute on the edge, meaning on location. Location could be a facility; it could be a truck; it could be a ship. It could be an offshore mine, offshore windmill, different locations as well an edge version of our database. And what is that? It basically is the same database that Christian has just described, but you can manage it through the cloud. So it's not an on-prem thing. It's manageable through the cloud, but the calculation, the analysis, and the capture of the data is at the location and all these different areas I just mentioned.
However, it's not just that the data is captured there, and it's stuck there. There is the ability to sort and to send specific data back up to the cloud just you don’t have to send everything up there. Think of temperature measure; it always measures 21 degrees. How much of those 21 degrees measures do you want to send up is really not that insightful. What you want to insight and send to the cloud is the interesting data, the anomalies, the summaries, the trends, those kinds of things. And what we see is that the learning models typically are in the cloud. And the analysis is then on location, taking care of the massive amounts of data that happens in a location. And so the combination of not only the remote management of what to do on the edge in terms of the software you have there like our database but also the selective sending and sharing of data between one site to a global office or to learning models. I think this is what the modern edge technologies will do for us. And I think that's quite fully in alignment with what our strategy is as well as Microsoft.
DAVID: It absolutely is. And that was a great trip through how -- Well, I learned a little bit there about how we only report back…I'll call it an anomaly or differential data as opposed to all of the data that you might collect on an edge device. And you're right; it does dovetail nicely into some of the recent IoT announcements coming out of Microsoft. And one of those announcements, in particular, was about the public preview of something called Azure Percept. And Diego, I wonder if you could tell us a little bit about Percept.
DIEGO: Sure, David. So Azure Percept, which is currently in preview, is a turnkey if you will edge intelligence solution. So it's a family of hardware, software, and services that are designed to accelerate computer vision applications. So we say that it's a [inaudible 29:30] kit with a camera-enabled, and it's deeply integrated with other Azure services like Azure AI and IoT Edge services. And it comes with a library of pre-built AI models for common vision scenarios like object detection, shelf analytics, vehicle analytics, and even some audio capabilities like voice control and anomaly detection. So basically, it's a family of products, hardware, and services that are bundled together to speed up the development of computer vision solutions.
DAVID: I'd be remiss if I didn't mention that there's also what we call Percept DK or Development Kit by ASUS with a carrier board, mounting tools, and camera-enabled system-on-module for the device. This is available now for people to get started developing with Percept today. I wonder if it might also be worth mentioning that we do have Azure SQL Edge, a very small footprint edge optimized SQL database that also supports AI scenarios. So we have made, or I should say, Microsoft has made a lot of investments here, and we're really doubling down on the future of IoT Edge. So moving just a little bit, and now we're seeking out your industry pundit knowledge and looking ahead to the future. What is the future of machine data, Christian?
CHRISTIAN: Well, it's going to be very exciting, and we're only at the beginning there. So, as we speak, a whole ecosystem around data is emerging: using data, storing data, selling, and leveraging data. The machine data use cases everywhere are simply exploding. The requirements to essentially manage the shape and the scale of such data, as we really discussed earlier, will hit almost any company in the world. Nobody can afford to let the data slip through the fingers like sand anymore. And with this growth, also the total cost will play an important role. And I think the combination of Azure as a global leader in operating compute and storage plus the power of purpose-built ISV solution like CrateDB combining that is serving exactly that need. And of course, in addition, as we have heard already from Eva, edge will gain significant importance and not just for the data collection and the buffering of data but really doing workloads on the edge with edge analytics and specific edge functionalities like Diego just described with AI and all of that ideally managed, backed up, and synced with the cloud. But being able to operate independently from the internet, I think that's where things are moving and what's going to be really exciting.
DAVID: That's great. And I wonder, Diego, if you could lend the Microsoft point of view to that question, and where do we see machine data technologies moving in the next several years?
DIEGO: At Microsoft, we refer to the intelligent edge piling on the topic of edge computing. The pattern has always been, or the motto has always been, "train in the cloud, execute on the edge." But now, as the edge devices are becoming more powerful and we are becoming smarter at packing more into smaller devices, we'll probably start seeing more training of AI models on the edge itself. So that unlocks much more interesting scenarios where the data is processed, and the models are trained on the edge, and then it’s sent to the cloud for other types of workloads. So it's evolving the concept of the intelligent edge, I would say.
DAVID: So just building on that a little bit, speaking of training AI models on the fly, on the edge, that means that they're constantly being updated with new data. Is that correct?
DIEGO: They can be. You're spot on. You can have more frequent training of the data and particularly in scenarios where the nature of the data changes. Often, you need to refresh the models more frequently. So if you can do that on the edge, it unlocks that. It makes it easier to make it more frequently.
DAVID: Wow. That's a really compelling technology. So it's fun to see where AI is going in addition to edge computing. So, let's turn just a little bit to how Crate.io and Microsoft have been working together. And I wonder maybe Eva if you could explain to us the experience that you've had working with Microsoft and how that's been for you.
EVA: Sure. It's been a very good experience, but we also realize we're a small company because we're a best-of-breed database company and Microsoft is a big company. So to know how to work with you is not that easy, but we're working on it. And my background obviously being an ex-Microsofty helps greatly there navigating the right teams. But basically, we work with you extremely well because clearly, Microsoft’s strategy is partner-focused and partner-centric. And also, we just highlighted where our sweet spot is and that there are surrounding technologies that have similar things, but not quite. So we welcome as a partner to be deeply aligned, especially on the technology side, which is the core. So we’re always in lockstep, and we're clear of -- We’re a company like [inaudible 35:53] provides the extra value versus the offerings that Microsoft provides as you move forward in the progression of technologies that they always stay aligned and we're not having a confusion here but also as you come up with new versions that we’re also aligned with you so that collaboration works really well. We are at the Marketplace, and the salespeople at Microsoft are also compensated by referring us, so that works well. And then, on top of that, we’re aligned in terms of go-to-markets in numerous countries. So I would say while Microsoft is a very large company, we have a global partner manager that helps us greatly, small plug here for Zarina Esimova, who is based in Austria because we are sitting in Austria. But she's doing a wonderful job as the global partner manager and helping us navigate the rather complex structures in this, basically you as a large company and a partner. But in summary, this is going extremely well for both parties.
DAVID: Well, that's great to hear. And thanks for that trip through what partnership looks like for you. I'm afraid we're getting down to the end of the show here. And I'll mention to the listeners that, of course, as usual, we'll include links in our show notes to the various companies and technologies and things that were mentioned throughout the show and to everyone's social presence so that you can follow them and learn more. And with that, Christian and Eva, thank you both so very much for being on the show today. It's been a great conversation.
EVA: Thank you so much for having us.
CHRISTIAN: Yeah. Thank you.
DIEGO: Thank you, Eva. Thank you, Christian.
DAVID: Thank you for joining us for this episode of the Azure for Executives podcast. We love hearing from you. And if you have suggestions for topics, questions about issues discussed on the show, or other feedback, contact the show host, David Starr or Paul Maher through the social media links included in the show notes for each episode. We look forward to hearing from you.