George Mathew is the chairman and CEO of Kespry, the leading drone-based aerial intelligence company, based in Menlo Park, CA. His focus is on leading the company’s mission to transform how people capture, analyze, and share insights about their businesses. Mathew has 20 years of experience developing high-growth technology startups. Prior to Kespry, Mathew held senior leadership positions at Alteryx, salesforce.com, and SAP. He has driven company strategy, led product management and development, and built sales and marketing teams. Mathew holds a Bachelor of Science in Neurobiology from Cornell University and a Masters in Business Administration from Duke University, where he was a Fuqua Scholar.
Amit Chaturvedy joined Cisco Investments in 2014 and leads the global investment practice for Internet of Things (IoT) and Enterprise Networking.
Amit Chaturvedy (AC): Tell us a little bit about what’s changed at Kespry since our original investment in 2017 and the progress you’ve made since then in the last two years or so.
George Mathew (GM): One of the things that I’ve noticed particularly in the industrial drone market is that we’re starting to see a fair amount of specialization. Providers are delivering more hardware solutions and more integrated vertical applications into the market. And this is where Kespry has continued to evolve since the investment that Cisco Investments made. We’re focusing more on software, specifically artificial intelligence-based analytics and applications.
AC: Makes a lot of sense. So, in your own experience, what have you seen as some of the things that have been going well from a partnership perspective and where you benefited from the Cisco value add?
GM: I think Cisco has traditionally taken a very strong leadership position in the IoT space. And we at Kespry thought we could focus and deliver more market value in this area.
And for us, it initially started from a relationship beyond the investment to really help us get into a lot of the key accounts that Cisco has. Cisco has strong customer relationships in our focus verticals and has made several introductions for us into those customers.
But that’s actually even gone further. As we start to get into more of the video and image processing side of where Kespry is headed, I believe that there’s even a more tighter relationship to be built with Cisco.
AC: And I can definitely vouch for that. You guys are a popular ask at CXCs/EBCs every time we engage with our customers and they’re looking to get access to innovation, and they’re very interested in how drones are getting utilized, and how people are combining drones with data analytics. Kespry sits right at the intersection of both those trends, so there’s always a high degree of curiosity.
GM: Just recently the Vale CTO and the entire team was visiting Cisco. Right in that first conversation, where our head of sales was able to connect with the Vale leadership team, the immediate output was a four-system pilot for Kespry in Brazil.
And so there’s a direct linkage between just being able to have those executive-level conversations at the right industrial companies that Cisco already has very strong account control over, and introducing a portfolio company like us into the mix. So we’ve been super appreciative of that opportunity to work with Cisco.
AC: Right. So it’s at that intersection where Kespry and Cisco together can repeat and rinse some of the conversations we’ve started with certain accounts, and do more of that over time.
GM: Yeah. It is interesting to see this. Because if you look at even back in 2015 when we saw a focus around big data analytics. The real moment where that market started to take off was to your point, Amit, when you started to see both Line of Business and IT leaders coming together, to drive initial digital transformation initiatives around their data. And now we’re seeing that in broader large-scale digital processes, extending in to physical assets.
And this is where it’s been such a natural fit to see how much a product like Kespry can be introduced even inside of an IT organization for the companies that we serve. The CIO and CTO see immediate value in this being part of the overall digital transformation of their next half a decade or decade to justify investment into their future.
AC: Right. So we’ve had a brief discussion about the go-to-market and the buying center part of your business. But, can you speak to the growth and maturity of the drone market itself? Because that’s the underlying hardware that you guys are using. And how has the industry landscape evolved? And how has that influenced your thinking of how you guys want to evolve your business over time?
GM: I go back to my initial point. We have started to see a lot more focus and a lot more specialization in the market. Kespry started in 2013, first creating autopilot and that autopilot solution had no hardware or airframe around it to be applied to industrial use cases.
So we literally built the airframe. And then we went on and built a cloud-scale infrastructure on AWS to be able to take the data that’s coming off of image and video sources on the drone and create the machine learning-based insights that are now driving the Kespry cloud.
And so, you move forward to 2016 and you start to see more focus come into the market. And we made an explicit choice to say our focus was going to be on the analytical data processing, machine learning and the software experience for our users. At that same moment, we started to see a lot of the capabilities on the hardware side evolve. Today we have a partnership with DJI. We take their drones and use those for industrial use cases.
And it’s a huge benefit for our customers because they can leverage the scale of DJI’s production. Our customers say it’s like the best of both worlds. You get the best of the focus and specialization on the software and the analytics, which Kespry is providing, and you get the best capability from a hardware perspective, which DJI is delivering.
AC: What problems are customers trying to solve? What outcomes are they looking for? What are some real-world live examples within your broader set of customers that you can share with us?
GM: You really have to have a real strong understanding of the physical condition of assets as they change specifically when corrosion occurs, when specific weather events occur.
And this is where Kespry has been able to deliver an enormously valuable solution for, initially the mining sector, now going into residential inspection, commercial inspection, and expanding further into the industrial inspection with the oil and gas customers we serve.
And really, that’s the only four segments that we target predominantly because we can help our customers understand how to take the data that’s describing the physical condition of the asset and marry that, for instance, with the telemetry that’s coming from those assets, the account for how much inventory production or asset maintenance value is related to that asset over time.
And today we do that for mining sites, for inventory of how much material is being brought out of that mine. We do that for insurance roofing sector, where we ensure the ability to have a very accurate dimensional analysis of a rooftop, and underlying structure of both residential and commercial property. And so, we serve a lot of the insurance carriers at scale for how they understand when damage occurs during a major weather event.
And then more recently we’ve started to marry a lot of telematic data that’s coming off of a pipe inside of a refinery – an undersea robotic rover collecting data for large scale pipelines that are being deployed in the North Sea. And being able to marry that data of video/ image analysis of the condition of that pipeline with the telematics data.
So this is where I see the future of a lot of these industrial IoT use cases evolve. Where this concept of perception-based data, being able to take image / video data for coordinate telematics, be able to understand where it’s placed in space, and time have a very strong analytical view, call it a digital twin of your physical assets is really what we think is driving digital transformation in the industrial verticals.
AC: What’s interesting to me is, all these use cases, while there is some commonality in terms of the digitization aspect and creation of the digital twin, and the need to deploy analytics to drive insights and outcomes, these are also very different use cases. An undersea robotic rover versus roof inspection, very different environments, very different buying centers, very different use cases. As a startup, how are able to scale up and service these different types of use cases? What’s the common base of AI technology, or machine learning technology, or the computing platform that you guys do that’s allowing you to scale up in these many different industry verticals?
GM: So what we’re realizing right now is there’s a whole new class of perception-based applications that are emerging because you have to get strong at really understanding image and video analysis. And this is where Kespry has been focused because of our heritage in building drone-based solutions for the last six years.
We’ve collected over a petabyte of information that’s regarding the video, the image of physical assets, and those physical assets and how they are changing over time, specifically with a lens towards geospatial coordination, so where it’s placed in physical space and how it changes over time.
And so, in all of those use cases, as right as is the want, the commonality is they’re very strong physical assets that you’re capturing high-resolution imagery, or high-resolution video, and being able to generate 2D and 3D models from it. And then, to your question on artificial intelligence, machine learning, there is a very common set of algorithmic methods that are being applied.
When you look at the world of artificial intelligence, machine learning specifically for this type of video analysis, it’s very clear that the deep learning models where you can actually train data sets of image/ video data to pick out anomalies like hail on a rooftop, or corrosion on a pipeline are being used at scale now to be able to, as I mentioned earlier, understand the physical condition of that asset and how it changes over time.
And so this is where we think that there’s an enormous opportunity that’s still underserved in the market. Take perception-based data around video and images of physical assets. Be able to apply machine learning in a way that enables you to understand how that asset condition changes, and really being able to produce a strong digital view of the asset performance over a period of time, which is really the holy grail for most industrial companies. To understand how well their assets are performing as they’re doing these industrial processes at scale.
AC: You also bring a pretty interesting point: while a lot of people are using standard AI models, you need real data to be able to train those models. And given the history of Kespry, by the time you started to more actively use AI and machine learning in delivering those solutions you were already a full-fledged business servicing these same customers, had access to all the data, and had a leg up versus competition in getting started on your own AI machine learning journey to deliver better outcomes for customers.
GM: Our journey because it started by generating the sensors, the payload, the hardware necessary to collect this data, and it was only about three and a half years. It was after I joined Kespry that we started to drive a lot of the insights, the machine learning, the artificial intelligence-driven insights from it. By that time, we had completed over 35,000 missions where we’re collecting data over industrial work sites.
Today, we’ve just crossed over 100,000 sites that we can now collect data for these industrial sectors. That is unprecedented. And that’s why we can actually now apply the AI/ML techniques that we’re talking about, because you have to have great high-resolution data to be able to do that work and Kespry is pretty much on the forefront of that workload.
AC: You had to go through your own journey of creating that relevance, that touchpoint with customers, and then transition as the market was transitioning, hand-hold the market and the customers through their own transition journey.
GM: I think you and I have actually talked about this in the past, and laughed a little bit about it. It takes several years to collect the right data before you can actually position yourself properly as an artificial intelligence, machine learning-driven company. Without that data, I don’t know how you can actually deliver any value from an ML, AI workload standpoint.
AC: Makes a lot of sense. So you’ve recently expanded your leadership team. Can you speak to the experiences that are shaping the culture at Kespry with infusion of new talent coming in, preparing the company, putting it on the roadmap where you and the rest of the board want to take it towards? How is the team dynamics evolving? And what’s the set of talent that you’re starting to bring into the team that’s different from before?
GM: Kespry’s crossed the magical $10 million ARR mark, which is a pretty interesting one for companies like ourselves. So we’re no longer a small startup but we’re not quite yet fully-scaled out as a business concern in this space. And so, there is a lot that we have to work through specifically when it comes to building the right team culture at that next point of growth.
So you kind of see these sort of moments of growth where initially, companies go into that sort of rocketship growth, the 0 to 10 phase, and we certainly saw some of that at Kespry. And then, when you start to get to the 10 million-plus you have to start to think about what that plateau looks like to build the next ascent, to be able to grow the next, call it, 15 million, 20 million in revenue.
And so, there is always this natural point in the enterprise software space. And I’ve seen this over and over again in companies that I’ve served with, where in that mid-teens moment you have to start to think about what is the right level of focus in terms of your go-to-market, your product-market fit, in terms of what you’re delivering to your customers, but also the team that drives that.
So, in that regard, we had some additional members join Kespry and we’re really excited about it. It’s been a pretty public information. First and foremost, Kirk Marple joined as our CTO, and he came in more recently to Kespry, about six months ago. And he joined with experience in two big areas that we care about.
One is being able to do large scale cloud transformation across the multi-cloud infrastructure. That was some of Kirk’s key background. And the other is being able to look at media asset management at Cloud scale. And so Kirk was one of the original developers that led windowsmedia.com at Microsoft.
Subsequently, he did a bunch of work for GM on autonomous vehicles, worked at STATS on sports analytics, no surprise on video image analysis. And so Kirk brought that level of focus and attention on where the future of Kespry is headed, particularly when it comes to perception-based analysis. And we’re really delighted to have a leader like him join as our CTO.
In addition, Langley Eide has joined as our CFO and Chief Operating Officer. Langley brings an amazing amount of leadership and solid understanding how to scale SaaS businesses. We had an opportunity to do that once before as we were at Alteryx together. I was the President and Chief Operating Officer of Alteryx, and she was the Head of Strategy and Finance.
And that was an enormously productive opportunity to define the entire self-service market together. Langley saw this opportunity to look at how perception-based analysis would change the entire industrial world. And she more recently joined us to help us do that next stage of the climb of the journey that we’re on.
So, it is something that’s important in every company’s lifecycle to continue to think about what is that right level of leadership to continue to serve the company at the stage of growth that it’s at. We’re at that stage of growth. We’re going from our teenage years to full-fledged adult years as the company grows, and we’re starting to bring in some of the people to help us get us there.
AC: Kirk and Langley are great additions to the team. We’ve been very impressed with the engagement and interaction we’ve seen happen. So, George, you have obviously the Alteryx background, you’ve been a senior leader in different companies like Salesforce, SAP as well. But yet you’re a first time CEO, right?
GM: Absolutely correct.
AC: So, I’m sure there have been several surprises, learning opportunities, failures that again became learning opportunities. So we’re looking for some practical advice that you can share with other entrepreneurs within the Cisco Investments portfolio. What advice would you have for other startup CEOs?
GM: So, there’s things that you come across as being a first-time CEO that you just don’t ever see until you are actually CEO. One, be very, very sure about what your product-market fit is, and be prepared to evolve the company, specifically when that product-market fit isn’t fully there.
And we saw that at Kespry. We started to see how the specialization and the focus in the market particularly between hardware and software evolved around drones. And we had to make very specific choices to say, “Hey,” who we were, call it for the next five years of the company versus the last five years. And that was the software focus we took.
The thing is, as you’re going through those changes, I would always recommend really keeping a very strong pulse on the culture of the company. And ultimately changes like the ones that we went through at Kespry to get to the point where we’re at were difficult. And so culture matters quite a bit because it keeps teams aligned and together to really focus on the journey in the long run versus all the short term wobbles, and weaves, and challenges that you run into as you’re going through that growth cycle.
Because as much as we’d all like to think that the journey is always up and to the right, it’s not. Often times, there are very specific roadblocks and barriers that you encounter, and it’s how management teams work around those things to create great companies.
The third thing I’d always recommend is just have a very strong board and advisors around you. We have been blessed at Kespry to have an incredibly strong board. Amit, you’ve been an enormously beneficial and helpful advisor to us from a board observer perspective. But as you know, the entire board is rowing in the right direction to help the company get to the next level of scale.
So, having those three things really well understood as you’re going through any entrepreneurial journey, I think is absolutely essential, particularly, one who’s either a first-time CEO or a first-time entrepreneur going through this. You’re not quite prepared until you’re actually doing it.
AC: Right. That’s really great and practical advice that I’m sure our entrepreneurial community in Cisco Investments is going to benefit from. So, thanks a lot for sharing that. And again, the journey is wide open, and the opportunity is wide open for Kespry. So really looking forward to what’s next and continuing to support you and the rest of the Kespry team in this journey.
GM: It’s been an incredible two and a half years for me. We’ve got the pleasure to know each other now for that same period of time, and we cannot be more appreciative of how much Cisco has enabled us to get to the point that we’ve gotten to.
As a sponsor of Glamour’s 2017 Women of the Year Summit, we launched a new mentor program in partnership with The Girl Project—Glamour Magazine’s philanthropic initiative. The Girl Project aims to unleash the vast economic and social power of girls through education to ensure that girls everywhere have access to quality secondary education. This mentorship…
In any sport, athletes and amateurs alike are concerned about how their equipment might impact their performance. For gamers, the capabilities of their hardware are fundamental to their experiences. It can be particularly frustrating when the viewing field of the game is interrupted by the bezels of three monitors, or by the slow response of…