In this episode, we discuss how manufacturers can save millions in CapEx and excess inventory by unlocking the shadow factory hidden in the operational inefficiencies. We also explore the value and complexity that comes with providing a full stack solution that integrates hardware, software and services. 

Our guest today is Saar Yoskivitz, CEO of Augury. Augury eliminates downtime and improves productivity with world leading machine health diagnostics.

IoT ONE is an IoT focused research and advisory firm. We provide research to enable you to grow in the digital age. Our services include market research, competitor information, customer research, market entry, partner scouting, and innovation programs. For more information, please visit iotone.com

Erik: Welcome to the Industrial IoT Spotlight, your number one spot for insight from industrial IoT thought leaders who are transforming businesses today with your host, Erik Walenza.

Welcome back to the Industrial IoT Spotlight podcast. I'm your host, Erik Walenza, CEO of IoT ONE, the consultancy that helps companies create value from data to accelerate growth. And our guest today is Saar Yoskovitz, CEO of Augury. Augury eliminates downtime and improves productivity with world leading machine health diagnostics.

In this talk, we discussed how manufacturers can save millions in CapEx and excess inventory by unlocking the shadow factory hidden in their operational inefficiencies. We also explored the value and complexity that comes with providing a full stack solution that integrates hardware, software and services.

If you find these conversations valuable, please leave us a comment and a five-star review. And if you'd like to share your company's story or recommend a speaker, please email us at team@IoTone.com. Finally, if you have an IoT research, strategy or training initiative that you'd like to discuss, you can email me directly at erik.walenza@IoTone.com. Thank you.

Saar, thank you so much for taking time to join us on the podcast today.

Saar: Thanks for having me.

Erik: So I always love to speak with founders and CEOs because there's always some interesting backstory. So it's not an easy thing, to do to quit a well-paying job and dump a lot of time and often at least lost income if not personal savings into a new company. Maybe this is a good place to start. Can you just walk us through the path that took you to set up Augury, I think about now about 10 or 11 years ago?

Saar: So we have two cofounders at Augury, myself and my partner. And we actually met during university. So both of us are from Israel, we have an engineering background, we started in the Technion, which is similar to MIT, it's a tech-focused university. And we spent five years, just talking about different ideas and businesses and opportunities. And it was clear to us that someday we'll start our own business, we'll kind of partner up and create something that provides impact to the world.

Specifically, at the time, I was studying speech recognition using machine learning and AI techniques. What we do today is very similar to speech recognition. We listen to machines and based on the noise or the sound, we can tell you what's wrong with them. We take an audio wave, and instead of searching for words, we search for patterns that indicate different malfunctions.

So the core of the technology came from there. And both of us had background in our military service in Israel.  We had background working with large machinery. Gaul, my cofounder was in the Navy. I was an artillery force. When your life depends on a big machine, you become very intimate with it. So you learn how to recognize every squeak and every sound that it makes. And it was pretty natural for us to take this into the understanding that we can diagnose machines based on the sound and the noise that they make.

And then we asked ourselves, okay, what type of machines do we want to diagnose? So we started talking to different markets. We talk to commercial buildings. We talk to factories. We talk to automotive car fleets. We talked to overseas shipping, until we found out this whole world of predictive maintenance that has been in existence for 20 years back then. And we found a huge gap. So the same techniques and methods they were using in predictive maintenance haven't changed since the 90s. And we felt at the time, it was 2011, we already had the iPhone. I think Alexa emerge or Google Home, we started to see the insurgence of AI for speech recognition. And we said maybe we can marry the world of internet age technologies with the industrial market, and really revolutionize or democratize predictive maintenance.

Erik: So maybe three of the probably the top five use cases for IoT are going to be machine envision, predictive maintenance and speech recognition. How do those differ in terms of the algorithms or the way you might approach them as a solution developer?

Saar: The chicken and egg problem of any AI company is that without data, you cannot develop algorithms; without algorithms, you don't have any value to provide to customers; and without customers, you don't have data. So it's like this vicious cycle. And you need to break free of it. Now, when you look at speech recognition, you can go online, you can scrape the internet, you can look at movies. There are different sources of data that you can create.

When you go into our market, this data set never existed. So yes, predictive maintenance has been around for 20 years, but it was siloed on desktop computers. No one really aggregated all this data. So we had to create our own dataset. Fast forward to today, we have over 150 million hours of machines that we've monitored, over 90,000 machines we've diagnosed and all this data sits in our cloud in our servers, which enables us to continuously refine the algorithms and improve the accuracy or create new types of insights on it.

So the first challenge was when we started was actually how do we get all this data. So that's one big difference between image base where you have tons of data sets for image recognition and maybe speech recognition when you have huge datasets versus when you get to the more niche problems that you're trying to solve, it's just very hard to get the data to begin with.

Erik: Even if you have this in the cloud, I think it can still be challenging? Because customers, do they try to silo the data? Do they allow you to kind of cross analyze datasets? And I mean, obviously, there's a big value proposition that's kind of everybody wins by contributing some fraction of insight to this algorithm. But nonetheless, people are still quite sensitive about their machine data.

Saar: The customer owns all the data, it's their data. But we get irrevocable license to use the raw anonymized data for our own research purposes. And in the beginning, it wasn't that clear for people why they should do it. But I think over time, they started kind of seeing the benefits emerge.

We had a conversation with the chief supply chain Officer of Colgate sometime last year and he told us, I don't want to win over my competitors because I'm better at detecting Berenguer. I want to win because I have a better product, understand my customers and better at marketing maybe. But if there is an opportunity to have societal impact, reduce energy consumption, reduce water consumption, and by doing so I can have rising tide lifts all boats, then the industry should come together, the industry should pull in all the data in order to provide that massive gains on environmental sustainability.

Erik: I remember, this is also about 10, maybe 12 years ago when I was doing my MBA and there was a case study which was actually Pepsi and Coca Cola. It was some type of algorithm deployed on production equipment to improve efficiency, and they would not agree to this. And it's the case study was basically like game theory problem. There's a win-win but both companies prefer the suboptimal solution because they have some concerns around this. So it's great to see that there's been somewhat of a mindset shift in the decades since.

Saar: There has. And also, we work with vibration data, magnetic data, temperatures, the more mechanical data set. Today, we don't rely on the operational data center, and that's typically where companies will consider their secret sauce. So to begin with, the data we work on is less confidential and they're easier to share.

Erik: Let's start with the value proposition. And maybe we can begin with a concept called the shadow factory, which we're discussing by email. So just outline this for us and explain why this is something that manufacturers should care about.

Saar: I mentioned we started 10 years ago as a predictive maintenance company. And as we had more of these conversations with everyone from the plant level to know the corporate executives, over time, we figured out that predictive maintenance or machine health rather, is not a maintenance problem, it's a supply chain risk management problem. So it goes way beyond just maintenance. That's why we also started using the monitory machine health and kind of are moving away from just maintenance.

And today, companies operate under an assumption that machines just fail surprisingly randomly. Once you ask them, hey, what would you do differently if machines stopped failing? How would you change your operations? You see these ripple effects. One of them is what we call Shadow Factory. So the largest cement manufacturer in North America expects that we will help them increase capacity that is equivalent to two full factories.

So you could imagine the capex investment that is needed to build two cement plants, or it's roughly a billion dollars. But more importantly, I would argue, you can look at the environmental impact of building two cement factories. And the challenge of hiring hundreds of people: talent is a huge problem these days in the manufacturing industry.

So hiring the talent and the CapEx and the environmental targets that they have, we're not talking about, hey, I helped you detect Berenguer, or helps you detect crack taller bars and the motor, you're talking about how can I help you reduce CapEx by a billion dollars? So that's the notion of shadow factories, full factory that are just hidden inside inefficiencies today. And if you could just work smarter and apply different technologies and digital tools to help you be more productive, you can basically increase capacity. And we've seen this over and over with, I would say most, if not all of our customers.

Erik: How do you do that? A lot of technology companies do focus on the KPI and say we help you improve some metric for this piece of equipment. And, like you said, it's probably not a very smart way to sell because then you are building a business case around maintenance budgets. But to move beyond that, you have to understand how improving performance in specific aspects of a factory can help unlock additional production capacity, which means also impacting forecasting. And the group of stakeholders that are somehow involved in making decisions around this becomes much, much broader because you're dealing with management level decisions around how much we produce at this factory as opposed to dealing with a production unit that's responsible for maintaining some pieces of equipment.

So how do you approach this then as a business? Who are you engaging with? And how do you broaden your perspective beyond the equipment that you're touching with your algorithms to the operation of the factory as a whole?

Saar: We're not here to sell technology. We're not a sensor company. We're not an AI company. We're a machine health company. So we come in to solve a very specific use case that I would argue all of everyone who owns a factory today is struggling with. And that is, again, the fact that machines just fail unexpectedly.

And we always search for that, we go in with that use case to begin with. And then we start with a very basic foundational level of let's make sure all your machines work, let's put this on all your factories on all your critical machinery and let's make sure that they just stopped bleeding with unplanned downtime.

After that happens, now we can go into that second order and say, hey, if machine stopped failings, why do you have $50 million worth of captured spare parts inventory sitting around collecting dust just waiting for something to fail? Maybe if we can commit to giving you a two week lead time on any malfunction, maybe you can just have one regional warehouse where you can ship the spare parts overnight? So that's one second order impact.

Why do you keep so much product inventory because machines tend to fail or need to be ready in the case of supply chain disruptions or whatnot? So well, if we can help decrease the risk of machine failure by 80%, maybe you don't need as much inventory, maybe you can become leaner, maybe we can enable this just in time philosophy that you're shooting for. So we start with the very basic level let's prove our worth, let's get very quick time to value. And today within a month all of our customers get full payback of the whole annual program and then we build from there.

Erik: Are you building this kind of one category of equipment to like say we start with CNC machines and we add categories organically based on customer demand or is the algorithm flexible enough that you could deploy it on an entirely new type of machine that you haven't worked with previously and it would learn how to function within a relatively short period of time?

Saar: So today we're focused on the process manufacturing industry. You mentioned CNC as an example, on the discrete side where we don't support or we're not active today. But unprocessed manufacturing lines, we cover 80 plus percent of the most critical machines, regardless of the industry. So, we have CPG and food and beverage and pharmaceuticals. On one hand, we have cement manufacturers and building materials and metal and plastic manufacturers. So we're very wide in our ability to cover.

And we only have one type of sensor or one type of hardware that includes vibration, temperature and magnetic data, that enables us to provide a holistic view of electromechanical system. Now the algorithms may change a little bit depending on the type of machine. If you're looking at the pump, you're solving for very specific malfunctions that happen on a pump versus a compressor versus a larger servo motor driven machine. So the algorithms may change, but the hardware, the basic sensing elements are similar. Out of the box, we cover 80% of the critical machines, and then that other 20% we can build over time.

Erik: You're working with process manufacturers, what about the equipment OEMs themselves? Do you have direct business relationships with them? Is it more of a partnership? Are they customers, how do you engage with those companies?

Saar: So the answer is yes. We work both with the end user, let's call them Colgate and Pepsi that you've mentioned. And then we also work with manufacturers of the machines that go on the production lines or in facilities in general. So today, we're partnered with Grundfos, which is the largest pump manufacturer in the world, as an example and carrier to the largest chiller cooling company in the world as well. I would say there's a larger trend in the industry of getting to outcome based services or water as a service as an example or cooling as a service or whatnot, we're kind of helping usher that change.

What you find out is if the OEM starts charging per gallon of water, instead of selling a pump, and the pump doesn't work, then they're not getting paid. So the risk of downtime just shifted from the end customer to the OEM. And obviously, as I mentioned, our system helps reduce that risk by 80%. So we're helping them build the backbone of technology, how do you connect a monitor of every machine? And our vision eventually is we'll have embedded machine health and every machine that goes off their production line.

Erik: You've already given us a bit of overview. But can you just paint us a quick picture of where you're coming from where you are today, and where you would like to be in maybe 3-5 years? Because I think it's always interesting to understand how tech stacks that have some aspects, which are exponential growth in terms of maybe the volume of data you have or processing capability, and then other maybe aspects of the tech stack that are improving at some single or double digit percentage per year in terms of effectiveness. How does that change look like? So it sounds to me like you've had a fairly similar tech solution for about 10 years now. But I imagine the efficacy has evolved significantly.  Like where are you today? And then if you fast forward, maybe three years or five years in the future, where do you expect to be?

Saar: So when we started augury we intentionally made the decision that we want to be a full stack solution from hardware all the way to AI and also professional services as needed. And today, 10 years later, we're still on that track. And being a full stack company has a number of benefits. First of all, it's very hard. Building a company that does both hardware and software, it's going to have to clashing DNAs and cultures and making them work and mesh it is really hard and we build it from the ground up so we became pretty good at it.

There is a flywheel of innovation that happens when you're a full stack company. You can look at other markets as an example, Apple, iPhone versus Android. But the reason that the iPhone scrolls so much better is because they control everything from the CPU where they create specific modules that enables their software developers to specific functions that delivers this hardware. Whereas if you need to build a general purpose solution that works with all these different skews and systems, it's much harder to get that level of fine tuning.

And in our case, our first product was a handheld device that connected a portable diagnostic tool that connected to an iPhone or an Android smartphone. And then a technician would go to a machine, attach the sensor with the magnet, collect the data, the data was sent to our cloud, we diagnose it and return the results to the smartphone. That was our first solution.

After we got to 50,000, 60,000 recordings in our algorithms, as I mentioned, our biggest challenge in the beginning was creating that data set, so that's how we build the dataset. After we had those 50,000, 60,000 machines that we recorded and diagnosed, we compiled all of that knowledge into our second generation of hardware which is a continuous diagnostics solution. The sensors, what we call halo today, where you physically attached the sensors to the machine, connect them to the cloud, and they send data in a continuous and we diagnose it.

Now we have 150 million hours of machine data and we're compiling all of this into generation or third generation of hardware. So having our algo developers and our hardware developers sit in the same room is really powerful and enables us to continuously bring innovations to the market that provides this two to three year lead versus other sensor manufacturers or AI companies that don't have hardware.

Erik: And if you look forward into the future, any significant changes that you anticipate or is this more of a continuous improvement process?

Saar: On the machine health side, we're continuously working to provide more coverage. So more coverage means today we're on the critical machines, how do we go to smaller, less critical machines? And last year, we launched a halo for supporting equipment that enables us to go to smaller machines that are maybe less critical, more redundant and whatnot.

Increasing coverage also means let's go to machines maybe that are not rotating at all. So there are other types of systems in a factory or in the industry at large that are nonrotating equipment. And today, we can maybe tweak the hardware a little bit, tweak the algorithms and create a full stack solution for other types of systems within a factory. So that's kind of an under machine health side.

And Moore's law continues to march on in our domain and cost of sensing goes down by 90% every couple of years. So we expect the next generation of sensors to be smaller and cheaper and better power consumption. And eventually you get to a sticker that you put on the machine. We're not that far away from it, maybe five years.

Now, another thing we're looking at is how do we marry the machines, health and behavior to the quality of the product that the machine is manufacturing, to the throughput, to the yield, to the sustainability and energy consumption? So connecting between machine health and what we call process health is very powerful. And today, we're already seeing the lines blur a little bit between these two worlds. And to me, that's the exciting part.

Erik: Just probably a month ago, I was interviewing a company called Williot, which does exactly this, basically, sensors on the sticker using Bluetooth. They're using them for supply chain right now because they can basically track location and temperature and you can start to add a few more sensor parameters. But I would say there's certainly not going to be robust enough to collect the data you need.

On the business model side, also, there's been a lot of progress here. So you were already starting to discuss the topic of shifting from a I-sell-you-a-pump to I-sell-you-the-ability-to-pump a gallon of water per second or whatever that might be. And If you're in the position of being able to tell a manufacturer, we can save you a billion dollars of CapEx investment, that at least has the potential to be a much more attractive business model than we want to sell you $10,000 per production line per year or something, SaaS solution. How does that look today? Because I think this is still a very new concept that a lot of both tech companies and manufacturers are trying to wrap their minds around. Are you already having uptake in this business model, or where are you today here?

Saar: The notion of performance based business models is definitely there. I don't think the market is mature enough yet for it, and maybe we as an organization are not there yet, as well. But that is definitely where the market is going.

So, we decided to begin as a pure SaaS model. So we are purely OpEx, we do have hardware, but we just give it away for free and charge on a recurring basis per machine per year. So that's where we started, which is already very forward thinking, if you can imagine five years ago, six years ago, in the industrial market, there was a lot of pushback for this model. Even today, some companies would prefer a $3 million CapEx in a one-time fee versus an ongoing SaaS based recurring model.

But a couple of years ago, we launched what we call guaranteed diagnostics, where we partnered with Munich Re, which is the largest reinsurance company in the world. And it took us five years of working together to get the confidence of their underwriting team in the accuracy of our system in order for them to back our diagnostics with their own financial products.

So today, if Augury says the machine is working well, and for whatever reason it fails, we will cover the replacement cost of the machine, both the parts and the labor. So this is the first time we're seeing insurance backed AI diagnostics solutions in the world, not just in the industrial market. So if you look at the healthcare, we're seeing more AI diagnostic solutions and they're not insurance backed yet.

So it was really groundbreaking a year and a half ago when we launched it. And we're continuously working on how do we refine it how do we get to the holy grail of manufacturing, which is uptime assurance? Not only do I cover the cost of replacing the pump, I also cover the production losses due to a downtime event. And when we get to uptime assurance, then it starts looking like a performance based model. But it's still driven through a very specific use case and very specific calculation of a value.

Erik: But if you have sufficient accuracy in the algorithm, and you're spread across a large enough set of customers, then you could diffuse the risk and it starts to act like a new insurer airplanes and you insure other things that could be $100 billion hit. But as long as you're covering enough assets that are diversified, then you can build a model there. How does this work with Munich? They basically integrate a service fee for this insurance product into your fee that you're charging to customers, or what does that look like?

Saar: Yeah. So for the customer, it's streamlined, they only interface with Augury. And we on the back end of things have our own agreement with Munich Re that covers the risk there. So not to get too deep into insurance mechanics, it's supposed to be as simple as possible for the customer.

Erik: First of all, I I've heard of a couple other cases that people exploring this, but I think you guys are sitting at the front of the market here. If we can return for a moment to just the simpler question of why somebody might prefer a CapEx over SaaS, why do you think this is? Because if you look at a B2C market, I guess there are a lot of areas where we do also prefer CapEx to SaaS. In terms of hardware, at least we prefer CapEx to leasing. But when you talk about software, the market is pretty much gone to SaaS, nobody's going to pay Spotify 5 years or 10 years or a fee to own it permanently. So it the value proposition is very clearly superior. And it feels like in B2B, it's also in most cases, clearly superior. Do you think it's more just the production processes, the accounting processes and maybe tax credits? Is that what is preventing companies from adopting this model?

Saar: To generalize I would say that if you look at broad enterprise B2B, the SaaS markets of year over year continues to grow exponentially. So there's definitely openness and maybe companies prefer purchasing in SaaS. On the IT side, the OT, Operational Technology, are not as used to it. And that's how their budgets are structured and how their ERP system is structured and they allocate different.

But if you look at the maintenance budget of a manufacturing company, they have a CapEx bucket and they have an OpEx bucket. And the CapEx is larger than the OpEx because that's how they're used to working: I need to purchase a new machine at a million dollars versus get a service that is ongoing. So this is historical, and I expect it to change over time.

We chose to be a SaaS company for a host of different reasons. But the biggest one is that it aligns Augury success with a customer success. Because if I sell you $2 million worth of sensors, and then they just disappear, and you fail, then I'm happy, I did my job. I got my quota. But if we have a SaaS model, every year, you need to reselect us as your vendor. And if we're not doing a good job, you can just fire us then and there. So we're very much aligned that we need to continuously do a good job to make you successful and win the business every year. And in order to do that, we continue to provide services, we continue to build our product.

If you look at our platform today versus where it was a year ago, we have hundreds of engineers working day in day out on adding new features without increasing the price, continuously adding more and more value to the platform. And this is why having a SaaS model really aligns to both the long term strategy of our customers as well as the ongoing day to day service in both the product technology side as well as the people customer success side.

Erik: Let's do a case study. I think the Pepsi one would be great if you can walk us through that. Because it's, now going to be interesting to look at what the journey looks like from identification of a problem to scoping out the solution deploying it. And if you can start with who are the stakeholders here? So who are the buyers? Who are the influencers? And who are the users of the system? And then just walk us through that process.

Saar: That's a very good distinction between the buyers and the users. So we are proud of our ability to work both on the maintenance technician level, all the way up to the chief supply chain officer, because you need both in order to be successful. Corporate team can make a decision spend a ton of money, but if there's not the proper habit forming on the production line and acceptance of technology on the production line, then they're going to fail.

So we go bottom up and top down at the same time. We find the main decision maker to be the VP of manufacturing or VP of engineering solutions, depending on the company. And increasingly, we're also seeing the IT department play a bigger role. So we had a couple of global rollouts that have been funded through the CIOs budget. And this, I would say battle, but hopefully it's more of a partnership between IT and OT, where traditionally, who's going to set the vision for digital manufacturing or industry 4.0 or whatever you want to call it? Is it the practitioners that understand the challenges of operations and maintenance? Or is it the IT people that understand the challenges of security and data lakes? Who's going to set the vision of what this may look like?

And increasingly, we're seeing more and more and more of the IT team getting involved, which I believe is a great thing because they know how to build scalable programs and scale really globally, whereas the OT historically haven't had this footprint. So these are the main personas we work with today.

Erik: And then if we look at maybe just this case in particular, and you don't have to talk about specific individuals, but can you walk us through the selection process for Augury through deployment and what the outcome look like?

Saar: In the last three years, a year before COVID hit, we started seeing a change in the market and COVID just accelerated tremendously, where companies already have a strategy. They worked on their industry 4.0 strategy on the digitization strategy, and they come a bit more well baked to the conversation. We obviously get help them providing our view and what we're seeing working with other companies like them to help build it together and help them shift their vision and strategy. But they're a bit more better baked now than they were three years ago even.

We're seeing more RFPs as an example, Request for Proposals. So we're seeing more of that which is an indicator of industry maturity. We typically come in work with, again, the corporate executives, VP, manufacturing and reliability under him or her. In order to identify the right test facilities, we initially start with two or three factories as a proof of value. As I mentioned earlier, in under a month, they get full payback of the annual program. And we start expanding, going from those two or three facilities to 20 or 30 factories depending on the company. And is it regional or global?

Erik: Would it make sense to dive into a particular case like this PepsiCo labs and the Frito Lay case? Or is each case sufficiently similar that there's not too much to digest in terms of the…

Saar: So we have a number of examples that are publicly shared. Frito Lay, there is a video, you can maybe show them the notes of the podcast, where they share our journey going from a four factory rollout to 30 some factories. And the trigger point for them is that within the first three or four months, we were able to help them manufacture a million pounds of Cheetos and Doritos. And they expect as we rolled out across all their factories now that we increase capacity by four months.

Going back to the notion of shadow factories, when they get to 12 months of increased capacity, that's a full plant that they don't need to build. So that's PepsiCo, and we've been working with PepsiCo Labs, which is their innovation team that are just tremendously impactful, and they have a really good setup there.

Another example is Colgate, Colgate within the first four months, we help to avoid them a failure that would have cost them 2.8 million tubes of toothpaste. Now, obviously, they know internally how to turn the number of tubes into dollar signs and they had very, very fast ROI. And they claim that within three months in one factory, we paid back a full year in six factories. So that made the decision very easy for them to roll out across their full global portfolio of both Colgate and Pet Hills, which is organic pet food company. So we have multiple examples like this where we saved millions of dollars in that first engagement.

And what's very important as we go into these conversations, even before we actually sign any agreement, is to make sure that they understand what they're trying to solve for. What are the KPIs that they're looking to prove? Because if you don't know what you're looking for, the experiment is going to fail by definition. So we make sure that they have set KPIs and also they're set up to quantify the value.

Because the worst thing we can do is, six months go by and we help them avoid 10, 20 machine malfunctions, but they don't really know what it means. Is this machine critical? How many hours of downtime did we help you save? How many dollars per hour do you manufacture? Is it $10,000? Is it $100? Are you looking to focus less on capacity but more on reducing maintenance and repair costs? So what is the baseline we're measuring against? What was your M&R spending the last year? So really want to make sure that we have all of that setup before we even engage in the proof of value. And that really helps everyone make a faster decision.

Erik: So you mentioned that you're full stack and you also provide services. Is that one of the services that you provide? So if somebody doesn't have that data available, would you go through a six month process of helping them to acquire some historical data here? Or do you just maybe advise them through this, but this is an activity that they would do themselves or with a system integrator?

Saar: So typically, we work with more mature companies, Fortune 500 and global 1000s and they already have a CMMS platform that has been rolled out or they're rolling it out now as we speak. They have some of the data available. We do provide the foundational work of reliability. So doing the asset criticality, making sure we have the ROI calculator, so if something does happen, we know how to quickly quantify it. So yes, we have a very consultative approach in ourselves. And the goal, again, is to make the customer successful, because if they're not successful, we're not successful. So how do we have a better setup going into these proof of values?

Erik: There was a point that you mentioned that you were working with PepsiCo Labs, which is an innovation driver at PepsiCo, is this typical? Because I'm sitting here in Shanghai, and we work a lot with innovation centers, a lot of companies are trying to basically figure out how to localize their technology, often on the product side, but also on operations in China to some extent. And so they're setting up these labs. But I would say typically, the ones that I see are more focused on the revenue side of the business as opposed to operations. But are you seeing a trend here where these labs are also taking an active role in identifying transformative technology partners for operations?

Saar: 100%, yes. So when I talked earlier about the maturity of the industry, so we're seeing more strategies around Industry 4.0, specifically relating to the operation side of business, so not just the sales and marketing. And the gains, there are just huge. Especially now with supply chain disruption and being in the news on a weekly basis, companies have come to an understanding that they have to focus on it.

Just word of caution for other companies, we love working with innovation teams, I think the Pepsi labs, specifically is maybe better than others. Typically, if you don't have the buy in of the business, the VP manufacturing or supply chain or engineering, then the innovation team can't really move mountains for you. It has to be a partnership between the business and the innovation team in order to really be successful. So after you have this first proof of value and you have the manufacturing team with a lot of conviction wanting to move faster, then the innovation team can provide access to the right executive instead of the program and pour more fuel on the fire. But it's very hard for them to go from zero to 20%.

Erik: Yeah, they seem to function best is the spark, but there needs to be somebody else that's going to pick that up and scale it. So, just last point here, maybe there's some something else that you'd like to cover as well. I think Q3, Q4 of last year, maybe Q4, a nice e-round, it looks like it was something like 180 million just from what CrunchBase is telling me anyways, so, first of all, congratulations. I mean, I guess raising a lot of money shouldn't be the goal of a startup, but at least it's somehow a necessary step towards becoming really an impactful company globally. What does this mean for you operationally? Is this driving new product development? Are you focused on moving into new geographies? Or what are your growth objectives in the coming year or two?

Saar: As you said, we look at fundraising as an opportunity to pull in new partners. These partners could be institutional investors and have experienced in scaling businesses globally in one case. And specifically in this round, these partners are industrial giants that we want to go to market together with.

So the funding will enable us to, on one hand, aggressively invest in growth within the manufacturing space and going into new segments, going to new geographies, going mid-market, going downstream, a little bit of smaller companies. That's one thing. But what's really interesting about the funding round is that it was led by Baker Hughes. Baker Hughes is the largest energy solutions technology.

And when we look at Augury, we have two strategic motions. We have a vertical motion, which is growing within manufacturing, as I just mentioned. And then we have a horizontal motion, which is how do we take everything we've learned in manufacturing and bring it into new markets? So, Grundfos as an example, is taking us to the water market and carrier into the commercial facility market. Baker Hughes now is taking us into the energy sector into the oil and gas, the drilling and refining into power generation in the future into alternative energies like wind turbines and others.

And then we also had Standard Electric, who invested in as part of the round and there are multiple strategic avenues we're pursuing with them as well. So to us, this funding is exciting, more so because it opens up the energy sector for us on top of just being cash infusion that enables us to grow faster.

Erik: It sounds like you have a great selection and very carefully selected group of partners here. I think we've covered a lot of territory here. Is there anything that we haven't touched on that you'd like to cover?

Saar: No, I think we covered a lot of ground. It was a fun conversation.

Erik: Yeah. Well, again, thanks for taking the time to talk today, Saar.

Saar: My pleasure. Thanks for having me.

Erik: Thanks for tuning in to another edition of the IoT Spotlight podcast. If you find these conversations valuable, please leave us a comment and a five-star review. And if you'd like to share your company's story or recommend a speaker, please email us at team@IoTone.com. Finally, if you have an IoT research, strategy, or training initiative that you'd like to discuss, you can email me directly at erik.walenza@IoTone.com. Thank you.