Podcast #213 Long Tail Value

TRANSCRIPT

Ken: Good day, and welcome to episode 212 of our Momenta Digital Thread podcast series. Today, I'm greatly pleased to host Himagiri Mukkamala, former CEO of Pelion. Hima is currently the CEO of a healthcare AI startup in the mental health space. Previously, however, he was CEO of Pelion, enabling customers and partners to securely connect Edge IoT devices over multiple cellular networks globally. He also had a long tenure at General Electric in the industrial IoT and digital space, leading the platform Predix's Engineering and Product Management organizations. Before that, Hima led various engineering teams at Sybase, building mobile platforms and playing a significant role in the acquisition by SAP. Hima, welcome to our Digital Thread podcast.

[00:01:27]

Himagiri: Ken, thanks for having me. You missed a part about how I have been on a six-month sabbatical after Pelion, and that's because I got injured. I wish there were more opportunities, and I'm sure there were more, but I should have taken those opportunities - time to reboot and refresh. The way I think about it, my digital thread was all matted up, and this was a great way to fix it. But joking aside, good time with family and friends. Today is precisely the end of the six months, and that's the time I gave myself. It's a good start. I'm getting started with this podcast.

[00:02:07]

Ken: Excellent. I'm glad you were able to use that for that. I didn't mention the sabbatical; perhaps I'm jealous. I managed once to have the discipline between two big gigs to take a 'sabbatical,' it was one of the most relaxing times. When we put on our exec search hat at Momenta, we often advise people who have just left a more extended gig to say, "Don't push. Give yourself three to six months." Because you will find something and, ultimately, take the time and enjoy it because it's a gift the universe gives us occasionally. I'm happy that you did that, and I'm even more pleased that you used our podcast to jump back into the game again here. We call this the Digital Thread podcast, and of course, you already started to answer the question, which was great. What would you consider to be your digital thread?

[00:02:56]

Himagiri: Coming from the industrial world, it's a simple definition for me. If I look at the industrial side of it, it's all about the data sets that capture the journey of an asset from design to retirement. If you look at me as an asset, in a cosmic sense- I'm hoping my digital thread, while analyzed, would show this journey of learning, experimentation, and, hopefully, improvement. I look at the face of my leadership during the GE days versus Pelion- now, I've changed a lot in how I lead and operate. I've picked roles that helped me grow without thinking about how much I'm managing without consideration of the scope and get into entirely new domains and learned a lot. I remember having this conversation with an HR person about the feedback I received about being very unapproachable early in my management career, and it needed a lot of self-analyses to correct those behaviors. If I correlate with the industrial side, the big theme in my digital thread has been enabling learning through intelligence based on the data captured. To net it out, if you look at my journey, we've talked about this. I feel like I've gone backward in terms of the traditional- what you consider a conventional digital journey, where everyone goes from hardware to software. Still, I went from a pure software space to industrial to a company-like, at the basic physical silicon side, and even to the manufacturing at Penguin. That's how I look at my digital journey, thread, and think about that phrase.

[00:04:50]

Ken: You've given us several great hooks that we will talk about later because this software-to-hardware and, conversely, the more difficult hardware-to-software journey is certainly a strong point in what you've been able to show. When researching this, I had to appreciate your byline, so engineer to CEO and everything in between. I'd call that an apt description of many of our Digital Thread podcast guests; it's what I like to call the practitioner profile. Your time at GE Digital is emblematic of this trajectory. You rose from Senior Director of Software Engineering to SVP of Engineering and Technical Product Management over a five-year tenure, so that was a nice trajectory. GE, of course, in the day was and has been the pioneer of industrial IoT. Please tell us about your remit and what you're most proud of.

[00:05:44]

Himagiri: It was going back to that engineer and CEO in between, one of the things I did as an engineer and people look down upon- I shouldn't say 'look down,' but think of building an installer, if you've come from the old school ways of building products, the thing you run to install a product on your desktop or server. People say that it's beneath me as an engineer to write code for that, and that's one of the things I did, and it felt like you should do everything to experience the breadth of what it takes to be a good engineer or a good CEO. That's how I thought of that tagline, "No job is beneath you." That's how I think about how you grow. But going back to GE- in my memory of GE, where it started, it was like a short, small group. In the beginning, 50 to 60 sat on just one floor. That's when I started when the whole notion of GE Digital was just 50 to 60 people. I remember the last count when I had left was around 10,000 people. That organization grew from 50 to 60 to 10,000. It was an incredible journey as far as I am concerned. When I started- and not just started- my remit as I went through five years out there was to build this team, build this industrial IoT platform that was conceptualized, and make it a reality. People I got to work there, both from the platform side, the Predix side, and the business side, there's a lot of talent and experience. It was so much fun, especially coming from a pure software company like Sybase, so that was very exciting to me.

Returning to what I'm most proud of, we made a new category called industrial IoT. That is something I'm proud of. Unfortunately, it didn't manifest into reality at GE regarding how it got deployed, and there were a lot of factors. Hopefully, we'll touch on that. But I'm super proud of having to create that category of a new business and working with the industrial businesses to enable new outcomes. Also, I was personally at a phase in my journey where I met many new people and the relationships I built there, and I've hired many of them throughout the rest of my career. Those two are what I would consider my proudest memories from that time at GE.

[00:08:26]

Ken: It's certainly at the time; they were the thought leader. As you say, they defined this category of industrial IoT; remember, some of the early marketing campaigns, i.e., the power of 1% and 2%, which I thought is still a brilliant exercise in impact and the impact one can have even if they save 1% to 2% of energy costs and cycle times or things like that. The other you hit on is outcomes and the outcome-based view of your solutions. The solution orientation, one might say. That journey from 50 to 10,000 people, however, is a reminder of the danger of moving too fast, at least in the digitization of industry. If you could go back a decade to the start of your tenure at GE, knowing what you know now, what would you do differently?

[00:09:23]

Himagiri: When I started the next role at Arm, I had a cheat sheet of what not to do. There is a theory that first-mover advantage doesn't exist when a true product market fit exists. Although, as you described, we created a category called IIoT, industrial IoT, there were many things that we could have done differently that could have resulted in a different outcome. Although nothing is guaranteed, and there were a lot of variables- to net out the things I noted down when I started my next job on things I should consider- maybe some of them do differently, some of them put more in terms of my learnings. The first one I thought through was that platforms are hard to build and sell in the market. You've touched on outcomes. A horizontal platform is very opaque without these right use cases and outcomes. It gets bloated and bigger because you're trying to support multiple businesses and outcomes, and it doesn't solve one use case to a large extent. One big lesson I've learned is to choose the first outcome, choose the first solution you want to build, and make the platform good at it, instead of just building this huge horizontal technology that can evolve to support multiple outcomes. That's the big lesson I've learned in thinking more outcome-centric, thinking of one outcome versus supporting multiple. The second, maybe more important or equally important, is sometimes culture is hard to overcome than technology. I should have spent a lot of time on the digital side. I should have spent a lot of time understanding how traditional industrial businesses sell products, work with customers, and how they release products into the market.

It was very different from how we thought software should be released into the market or should help the customers. That culture of how traditional industrial businesses think and operate versus how the software businesses should operate- merging those two should have been done better from my and the team's perspectives. You've touched on this; we grew too fast, from 50 to a huge organization. We've learned smaller teams make a bigger impact as there was more visibility and focus on- because of the good marketing impact, the team became too big, and it became harder for us to move, and change based on the customer and the business requirements. That was a huge operational challenge, so that's the third learning. Maybe tied to this is the notion of incremental build versus a big platform. If you look back at our huge marketing push, it turned out to be this big platform that supports everything. If I had to redo it, I would think about doing this incremental way of building, like I said, with one outcome. The final lesson I learned- we can think of software in a very agile, incremental way, but customers don't think that way. The traditional industrial customers- large customers, long life cycles, and industrial assets last for 40 years, so how they think about software is very different from how we think about it. So, finding a way to merge these long-life cycles of the assets with the short life cycles of the software. We should have done something differently instead of confusing the customer base. Those are the five big lessons I've learned: people, culture, technology, and understanding the customer base better.

[00:13:34]

Ken: Those are five great learnings indicative of the difference between OT and IT. Think of it in terms of general traditional business software being IT, OT being control systems, and the upper intelligence layers building out that control. It's interesting because OT will never move as fast and dramatically as IT. After all, it's got a huge amount of accountability and responsibility in terms of the regulated systems and such and the long-life cycles of the assets that go with it. It's apropos, and we should do a whole podcast on those five points. Knowing the bio upfront, I mentioned Pelion. It may not be clear to the group because you've mentioned Arm several times. You went on to Arm and took a very ambitious role there, going from General Manager of IoT Cloud Services to the SVP of the Pelion IoT platform, which you will ultimately become CEO once you spin it out, a very ambitious move that ultimately became Pelion. What attracted you to Arm at the time, and what were some of your key accomplishments?

[00:14:43]

Himagiri: To reflect on the first topic, we talked about my digital thread and how I took some rules that are not tied to the scope or that enabled my learning. I went from a large organization at GE to just managing a few people. Because I wanted to step into this aspect of the product lifecycle, the hardware side, I stepped down from a size scope and went to a newer area. That's the transition from GE to Arm. The reason I took it- if I looked at my experiences at GE, it taught me- it's important to understand the whole product, not just one piece of it, the software. As I looked at IoT, enabling these outcomes is very important, especially from a security standpoint, to understand the stack down from the silicon to the software and how this integrated product gets delivered. If I looked at the two companies in the market that do that, they were Intel and Arm, but many sensors and the Edge nodes were being built on silicon. When Arm reached out to me and said, "Hey, do you want to work?" I asked, "What do you guys do in this space?" It was purely from Cloud and software and services. It sounded very interesting when we discussed it; it aligned with how I thought products should evolve. This notion of an integrated hardware/software stack being essential from an IoT space- there's no better place than Arm. That's what made me go to Arm. Again, I chose roles that had to be very different.

GE was a very different organization; GE sold through these industrial businesses with long attach rates with customers the sales team has worked with for a long time. If you know Arm, Arm doesn't have a direct sales team; they're selling through the ecosystem and the partnerships through companies like Silicon Labs that license their technology and sell it to their end customers. Arm doesn't do any manufacturing. That's something that attracted me in terms of learning new things. Also, the people side of it, Simon Segars, who is CEO of Arm, and Dipesh, who was CTO- they were very different leaders from a hierarchical organization like GE. They were very hands-on and approachable, so that's the other thing that attracted me a lot from- how it was different from GE, which I liked, so I went there. It was an interesting five years; for me, two years is never a good time to make an impact, so I stayed there for five years. It was quite the experience because- and we might touch more about this later, going to a hardware company and trying to make that team operate like a software company; for me, I'm proud of something- that is the biggest thing I'm proud of because the hardware has to be 100% perfect, right? No defects. But the software doesn't evolve that way, so I'm incredibly proud of trying to change the organization's mindset. Not just me but the whole team.

[00:18:13]

Ken: Yes, it's quite an accomplishment and certainly on a huge platform of Arm but also of Pelion. You presided over Pelion's spin-out and ultimate trade sale; both are defining achievements. What did you learn from this time about the IoT connectivity space and this spin-out sale process?

[00:18:31]

Himagiri: One of the biggest lessons, and one of the biggest things I was proud about, was three of the top five industrial companies- we brought them as customers to Pelion when we were inside Arm. But then the challenge was, how do you make sure going back to this theme of industrial companies, once they are a customer, they have long cycles, so you have to make sure you support them through their journey, a 20-year cycle and a 40-year cycle. When it came time to spin out Pelion from Arm, my biggest challenge was how do you take these industrial contracts, contracts with industrial customers, and ensure there is continuity, both from a support side, technology side, and the people side, as we go through this process of spinning out. Just as a background, we were a business unit within Arm, and through the soft bank and Arm process, we were spun out as a separate company, which has continued to run as a CEO. Like any other spin-out or acquisition, you think it takes X time. The reality is it will take at least 2X to 3X. I've seen that happen multiple times, so one of my biggest lessons was that as an engineer, we are told we always buffer time.

As a CEO, I should do that more, especially when you go through this acquisition process. As a big company, we had a complicated legal entity structure. Companies or companies and so on are trying to unravel all that and simplify it to a small business. Because at the end of the day, Pelion became a small company, which was a big challenge. I think there are a lot of new-age companies where you don't need to set up your legal entity to hire people in different regions, so at that time when we started, we didn't have that luxury. But hopefully, that doesn't become a challenge now. What else did I learn? Try to get a higher enterprise value, as the buyer will find ways to discount from the initial term sheet to the due diligence; there will be some gaps in what you provide. Always shoot for the highest enterprise value initially, assuming the eventual value you'll get will be discounted. That's something I've learned through that process, purely from a process and evaluation standpoint. But from a pure technology standpoint, the market we were in, this whole IoT connectivity space- it's just getting started. It predicts around a trillion connected devices by 2035; we are in 2023. In this market, where you talk a lot about Edge devices and sensor devices, that market is just getting started, so I'm still very bullish about it. I'm still looking at many more of these devices to come online and collect the data so that you can run AI workloads at the Edge as many of your investments do.

[00:21:50]

Ken: Adding low power wide areas like LoRa, Sigfox, etc., into the mix to help with those as well. It's become more diverse in terms of Edge communications, but we're still scratching the surface of what that looks like. I want to go back to a question you've asked a couple of times, and you and I discussed it at length before the call. That is this idea that we are companies trying to create software capabilities, and you have a long experience base and some passion. I was hoping you could say more about your own experience, and two, what you have seen in terms of success patterns there.

[00:22:30]

Himagiri: Ken, I wish I had a silver bullet for every company that wanted to do this. Fundamentally, when these companies try to add software, they must have a clear product strategy. There's always confusion about- we are adding a new product within the same company. Are we just enabling more hardware sales? Or is it a combination? Because that clear strategy enables a hardware company to become successful, and often, there isn't clarity to start with. I'm not saying you should stick with what you start with, but at least the steps you take must be driven by that good first decision on what role software plays in your company's future. That's one thing that should be clear from the get-go, and that will enable many downstream decisions in terms of pricing, go-to-market, and so on, so that is an important decision. The second thing that I've seen that companies can do earlier is a clear organizational strategy. As I've said, cultures are harder to overcome than technology, and most of the culture is in the organization. Trying to structure the organization aligned with the strategy so that- whether it's an integrated product offering or a pure software and pure hardware, whichever the strategy might be, an organization that aligns with it, I think, is one other key sort of decision that needs to be made when companies think about driving towards new software offerings, integrated or otherwise. The third thing is 'build versus buy' decisions. I've seen a lot of hardware companies who traditionally built a lot of hardware; you end up building a lot of stuff on your own, trying to build software too.

If you come from a traditional software space, you tend to go with the mindset that there's enough open source in the market; I'm going to pull along and only build what's needed. This early decision on build versus buy also comes from the right talent, having the right talent in the organization that can think that way, and so that's probably the third one in a build versus buy decision. The fourth one I can think of is merging these product life cycles. We talked about it in the GE, long hardware cycles versus short software cycles, it creates friction in the organization, so having that clarity and talking about it within the organization and having a release cadence that bridges these two, that's very critical in succeeding in this transition in this digital transformation for a hardware company. Finally, we talked about it in the beginning, the solution-centric versus the platform-centric. I strongly believe solution-centricity helps hardware companies get closer to the customer and align their software with how the company delivers capabilities into the market. Those are the five or six things I think of key parameters from evolution into a merged hardware/software company.

[00:26:04]

Ken: Certainly, speaking from deep experience, because you've been through this cycle a number of times. Industrial automation is at an important crossroads given reshoring, the energy transition, and labor challenges. What key trends are you watching relative to the industry's future these days?

[00:26:23]

Himagiri: You and I have been in the space for so long, and it's a great time to be in industrial automation, especially in the US, all the reshoring effects, the CHIPS Act, and the manufacturing stuff being broad. Although I read just yesterday that TSMC has pushed out its plant in Arizona from 2024 to 2025 because of a labor shortage. It's a good time to be in the space. Labor and talent are a premium, so I'm keen on and interested in that. One of the key trends I'm seeing is leveraging AI to mitigate some of the labor shortages. But I honestly feel like computer vision is one of the things that is underrated and has gotten the last little bit in the midst of this generative AI mania. I feel like- it's also AI, it may not be generative AI sort of paradigm, but in industrial automation, especially in quality and safety, it plays such a big role, and especially automating the manufacturing sites and using vision to do that- and given the labor shortage problems, I think computer vision is going to get bigger and bigger, especially using Edge sensors to capture the visual data through the manufacturing lines. That's one key thing that I see as a trend emerging. The second one I'm passionate about, and you and your team are passionate about, is Edge and sensors, whichever terminology you use. Small sensors, big sensors, light Edge, big Edge, small Edge- ultimately collecting data and analyzing it, it's going to become a big differentiator in this space, in industrial automation, not just in the manufacturing side of the lifecycle but also where the assets are deployed in the field. That's a trend that we've seen over the last ten years but given the growth in these sensors to run AI workloads and also the growth in connectivity, as you talked about, and the diversity of connectivity available, I think it's going to become a big differentiator.

Finally, on the theme of connectivity, without all of these data sets being sent to either the local sensors or to the Edge gateways or the Cloud, without connectivity, nothing can be materialized. I see diverse connectivity options continue to increase and the emergence of private LTE, secure high volume. That's one other trend that we were also looking at when we were at Pelion. But I look at these four trends: AI, vision, Edge, and sensor data to stream that data so that analysis can be done, and finally, a good connectivity backbone that enables data movement. That's what I see as emerging trends or continuing trends that will be solidified in this space.

[00:29:46]

Ken: I think you've just described a good third of our portfolio and probably some of where we are making the most investments, especially the computer vision. It's nice to have that reinforcement. I know that you are running mountain ultra-marathons when you are not leading industrial impact. These are somewhere between 150 miles, plus some are over 200-mile races, such great locations: Tahoe, Bigfoot, and Moab. Tell us a bit about this passion.

[00:30:18]

Himagiri: That also needs one complete podcast.

[00:30:23]

Ken: Yeah, on running techniques.

[00:30:25]

Himagiri: Simon Segars, CEO of Arm, left a note on my LinkedIn post when I said I was taking a sabbatical. He left a post about my stamina and persistence, and I can attribute those characteristics to this passion. Ultra running has taught me that the journey is what matters. If you look at these long races you've described, there are eventual highs despite the long lows that you might go through. But you have to wait for them. Our career journeys are similar. You have to have this sense of optimism and happiness, and if you don't have those when you're running through a race that goes over 240 miles, that's 4 to 5 days- you will never try these races, and you will never enjoy these races. That passion has taught me those characteristics I take back in real life whenever possible. The other thing- I just came back from Colorado, where I helped a few of my friends run a couple of these long races. The other thing that this passion of mine brings me is a sense of teamwork and shared success. I was there for their race, not running a race. Sometimes, happiness comes from success, which is not even your success- them finishing their races or giving their all for the race. This is what this passion of mine brings to me, which allows me to go to work and think of shared success and team success.

[00:32:07]

Ken: Awesome, well said. It's great to see the physical and personal manifestations that make the profession. I agree with you there; there are probably a lot of similarities in character perseverance, stamina, and cadence- all of those things you learn from long-distance running or any sports. In closing, I always like to ask where you find your inspiration? Of course, beyond running mountains and ultra-marathons?

[00:32:33]

Himagiri: I listen to many podcasts, and I feel a few books have fundamentally changed my outlook on life. One of the first books I read, maybe in my early thirties, was "The Monk Who Sold His Ferrari." It was a great refresher in rethinking life passions and what matters in life. It's a book by Robin Sharma. The second book I feel has changed how I do stuff is by Sakyong Mipham, if I remember correctly. The book's name is "Running with the Mind of Meditation." Now, as one of those people who are, like, meditation is not for me, you sit in a place and try to do- it's hard to do. But reading that book changed my outlook on meditating when you run and do whatever you do. It's just about being present and focusing on life wherever you are. That's a book that changed how I think about life and breathing a positive outlook on life. The last thing is probably more similar to people. This book is about the great power of passion and perseverance. I think it's by Angela Duckworth. It discusses how the effort will count more towards the outcome than just raw talent. There is this equation where she talks about the effort you make, the persistence, and the talent; that's what differentiates you. It's a great book on how you continue- it fits into how I run my races, and those are probably the key ones that I feel impacted- along with all the professional podcasts I listen to.

[00:34:29]

Ken: Great recommendations. I enjoyed reading "The Monk Who Sold His Ferrari" several years back. "Running with the Mind of Meditation" is one I will have to look at. It reminds me of Eckhart Tolle's "The Power of Now," at least how you described it, which is also a read one of our other podcast guests recommended. Of course, "Grit." It sounds like great reading material. Hima, thank you for sharing this time and insights with us today.

[00:34:57]

Himagiri: Thank you so much for all the people you bring on here and for enabling us to learn from their experiences, the shared experiences at work make us human, and I'm so glad you are doing this consistently week after week, or every two weeks- that requires a lot of passion and persistence, so thank you.

[00:35:20]

Ken: I appreciate the mention. It is an effort of love; I appreciate smart people and those operating in the same space that I've come to love over 40 years of my career, and you have been a leader in that. I look forward to the next big things you will accomplish, given that you've had such a great track record and a great running technique and pace. This has been Himagiri Mukkamala, former CEO of Pelion and a mountain ultra-marathon runner. Thank you for listening, and please join us for the next episode of our Digital Thread podcast series. We wish you a momentous day. You've been listening to the Momenta Digital Thread podcast series. We hope you've enjoyed the discussion, and as always, we welcome your comments and suggestions. Please check our website at momenta.one for archived versions of podcasts, as well as resources to help with your digital industry journey. Thank you for listening.

[The End]