Podcast #178 The Chief Product Officer

TRANSCRIPT 

Ken: Good day, and welcome to episode 178 of our Momenta Digital Thread podcast series. Today, I'm pleased to host Suhail Jiwani, Chief Product Officer of Kelvin AI, a company helping industrial customers to simply and quickly scale control applications that create sustainable business value. Before Kelvin, Suhail was Chief Product Officer at Honeywell. He oversaw its overall software portfolio for connected industries and launched new trailblazing applications in connected worker and remote industrial operations. Before that, Suhail spent over a decade building and leading innovative products in industrial and enterprise IoT mobility and analytics at Hitachi, Cisco, Samsung, IBM, and SAP, all great names. Suhail, welcome to our Digital Thread podcast today.

[00:01:34]

Suhail: Thanks, Ken, for having me on the podcast. I've been a big listener of the Digital Thread podcast for months, and I'm excited today.

[00:01:43]

Ken: Right. We're very excited to have you, and I think it caps off a great set of 100- and now 78 episodes that we've had- so looking forward to this conversation. Of course, you'll know as a listener; we always like to start by asking what you would consider your digital thread? In other words, the one or more thematic threads that define your digital industry journey.

[00:02:05]

Suhail: That's a great question. Everybody has a different story, and I enjoy hearing those stories. If you ask me about my digital thread, it has to start with- I would say, my first job. Straight out of college, I joined SAP. And in my career, you'll see that even from the start, my focus has been on building new innovative products. Thinking out of the box to look into new transformational capabilities and how technology can help move us to the next steps. So even at SAP, my team was called- I was part of the team focused on customer relationship management products. But my focus was my team was called CRM Innovation and Introduction Team.

The focus was on how we take the existing workflows of the CRM application and think about that in a new way to extend this to mobile phones. Brand new smartphones can allow salespeople to create leads, manage leads on their mobile phones- on a smartphone. We are talking about less than 100 enterprise applications on the App Store or Google store, for that matter, during that time. We are talking about use cases that can integrate into Facebook for loyalty management, with Twitter for automated service ticketing. That has been a bedrock for me to start thinking about how technology can impact the daily lives of our enterprise users. And that continued even after I left SAP to join IBM to work on big data. And then, Samsung focuses on the new aspects of bring-your-own-device, that BYOD strategy around how to leverage the concepts of mobile security and mobile device management and bake that into a consumer-friendly and enterprise-friendly product that can be adopted by customers worldwide. And this led me to learn more about the telemetry aspects of things, the sensory expert of things right now, which led me to join Cisco. And that was the first favor of me learning more about the IoT- the industrial IoT, the smart radio, the smart concepts of things. That was an excellent experience for me to learn about thematic threads and how this digital transformation is helping this age or industry that has been there for 100 plus years. It transformed itself by leveraging these new digital technologies. That journey continued through the Honeywell and the Hitachi and all the good names you mentioned. The work I'm doing in Kelvin is bringing in all that experience and knowledge and taking our customers on the journey of autonomous manufacturing, sustainable manufacturing in the future.

[00:05:05]

Ken: As I listened to the words used to describe- I see on one side technology innovation, and on the other side, you've mentioned clients, customers, users. And it's interesting to see you're at that intersection of the two. Thus, Chief Product Officer seems an apropos title for that. What stood out for me in all of your experiences is this leadership in product management- so perhaps, that's a theme we can explore today. Let me ask a guiding question. What does product management mean to you? And more importantly, what are the hallmarks of great product managers?

[00:05:38]

Suhail: That's a great question. And it's kind of a very personal question to me, even if I'm running the product management, engineering design teams today. In my heart, I consider myself a product manager. Product management is as much of an art as it is a science. It's the art of making decisions about something which does not exist today as a product or a feature in a product. However, I believe that to be successful in product management, there are a couple of key fundamental traits that every product manager must-have. The first one to me is customer empathy. As you mentioned, technology and customers- everything else is excellent. We need to understand why the customer has a specific problem and what they need versus what they want to succeed as a product manager. If you cannot understand your customer's needs, you will never be able to build a product that your customer will use and believe in. Customer empathy, to me, is one of the most fundamental traits to be successful as a product manager. The second aspect to me is decision-making. The decision-making has to be data-driven; it must be outside in by understanding the market, customer behavior, and competitive behavior in which industry and vertical you're playing. What is the differentiation that you're trying to create? There's a lot of data to be gathered. But at the same time, if you are building something new, the data may be limited. Suppose you're working on version 1.0 of a product which you always end up doing. If you're talking about transformation, it's about the data that you can gather as much as possible. But at the same time, using intuition based on your experience, you have to make the right decision. That combination of data and intuition is the art of product management. But I cannot say a product manager can be successful without stressing that you need to have strong communication and evangelism skills. It's very important as a product manager to clearly articulate the customer and market problems and the needs to your internal teams like engineering- who rely on you to provide them the data and the context behind your product strategy.

Communication to me is another key aspect of a successful product manager. And on the other hand, you also need to communicate your product strategy, product roadmap to your customers, sales, marketing. Your internal and external partners gather feedback and continuously calibrate your roadmap based on new data and information in an agile fashion. And lastly- this is where a lot of good product managers, even though they are following all the right steps, the three steps I mentioned right now- fall into this. I firmly believe that to be a successful software product management leader- you need to focus on getting the product out in the hands of the user as quickly as possible versus getting into this rabbit hole of creating a perfect product; this is version 1.0, which may take 18 months or 24 months to get it out in the market. I would focus on getting the wireframes, the mockups and MVP, a Minimum Viable Product, a minimum viable feature in the hands of the users early in the lifecycle to get the early feedback, calibrate, or pivot even, then wait for many months to get the product out in the hands of the user, only to see that my market advantage is gone or I focus on prioritizing the wrong features. These are some of the key traits I have learned over time that make a successful product manager.

[00:09:27]

Ken: Very good list. My summary here: customer-obsessed, data-driven, evangelistic, and pragmatic with a high sense of urgency in terms of the four. I know you've got a lot of great experience and many great companies, and I want to drill into Kelvin. But I've got to ask, in 2013, you joined Cisco overseeing the product management for their IoT cloud offering. And of course, for all of those that were part of that era, it was a significant time because Cisco was the Internet of Everything, a great marketing campaign. What were some of the key lessons you learned, especially relevant to this positioning?

[00:10:07]

Suhail: It's funny you remember that time about 2013. It's almost a decade now. And it's suddenly hit me while you were speaking, but I'm much older and more gray-haired now working in this area. So yes, you're right. This was a time of Internet of Everything, with over 25 billion IoT devices connected to the internet by this year, 2022. That prophecy that was made in 2013 is almost true. We see billions of devices already connected to the internet to grow to 50 plus billion devices in the next five years. Edge computing concepts during that time were fairly new, and my team worked towards launching the world's first Edge computing stack at Cisco. Before this recording, we talked about this whole hype of Edge and Fog computing and how it will eat away the revenue of Cloud computing, which was a bit of an exaggeration for me then. And now, I'm glad that we have now learned our lessons over the years through multiple POCs, POVs, and going through some production instances. Now we have learned that it's not that Edge will eat the Cloud or vice versa. They both have space, and Edge and Cloud computing, to me, are more complimentary. And if you have to scale, we are talking about this distributed computing concept and the open systems from Edge to Cloud connected. That is the right way to scale going forward.

But if I have to divide my time at Cisco, I will broadly divide that into two phases. My first focus at Cisco was on the data plane, bringing in some new insights and analytics, and building new products and businesses for Cisco. Just to let you know, about 50% of the world's data flows through the Cisco network. The first part of my job was to understand how Cisco could tap into this network data to provide, for example, user behavior analysis or analytics on business performance through machine learning and big data analytics. I'll give you a very small example. When you go to a Best Buy, even if you're not connected to the Best Buy Wi-Fi network, your smartphone is always thinking about the available Wi-Fi routers. Your smartphone has a unique ID which is called MAC address. The Best Buy or any retailer for that matter where you visit can easily tap into that network data to understand how many times this smartphone, hence you, came inside the store. I can understand which section of the store you were dwelling in through triangulation, which means you're spending more time there. If you are connected to my network, I can know if you opened up something like amazon.com in my store. That means you are doing some price match. Can I use all this data and send it to a store manager who can give you a real-time offer of 10% off on an appliance or a TV to convert you immediately as a customer? That's a huge value prop. That's a different way of thinking about converting a user into your customer by understanding their behaviors.

Similarly, we can know patterns based on the network data coming in from various sensors that are connected to the network to understand how and where people are spending time in the store, or maybe on the factory floor, or in a smart city or smart venues for that matter and give them up to various types of business and user analytics. So that was the first focus of that. The second area where I focused on, I would say, is more on the IoT network plane. I spent time leading the Core Platform Product Management for Jasper Wireless Division, which was Jasper wireless was an acquisition by Cisco. My focus was on IoT device connectivity, building, and subscription management for over 100 million-plus IoT devices globally. If I have to summarize, my biggest learning at Cisco has been the experience of correlation between the network and the data plane. That helped me personally in my future work, and that's the same kind of experience that I'm applying today to the work that I'm doing at Kelvin- bringing in information about the network, the data, and now, even the control systems together to close the loop.

[00:14:27]

Ken: That's a great way to weave the story together. I'm dying to jump into the Kelvin topic. Still, I have to stop you at one more point, and that is you had the top product management role at Honeywell in 2019- one of the granddaddies of industrial control. You were a Chief Product Officer for the Connected Industrials division. Tell me about your role, some of the key wins, and then jump into Kelvin.

[00:14:51]

Suhail: Sounds good. I would consider Honeywell to continue the work that I've been doing at Cisco around industrial IoT. And of course, after my work at Hitachi- my team built and launched the Lumada platform not just to transform Hitachi globally but also their internal divisions and the customers working with Hitachi right now. So that was a vast experience. It was a continuation of work at Honeywell, although at a larger scale from the responsibility point of view as a Chief Product Officer for the connected industrial division. But I would write my time at Honeywell as one of the toughest yet most rewarding so far. I'm talking about when we were hit by the pandemic three months into the new job in 2020. And it fundamentally changed our, as well as our customers, view on how to run the business and why digital technology adoption is key to navigating through multiple challenges that companies face during that time and honestly are still facing and going through those pains at this point right now. So, if I divided my job at Honeywell, I would say that there were two parts to the job. And the focus was there, and I would call out some of the successes there. The first part of the job was to transform the existing product portfolio to be more software and licensing driven rather than services and projects driven to avoid the cyclical nature of the business.

Getting hit by the pandemic also meant that many projects in the pipeline dried off. How can we avoid that cyclical nature? How can we diversify the company through the software and the digital approach beyond oil and gas, beyond just the key verticals that Honeywell has been focusing around, say, mining and chemicals and oil and gas only, and which had the maximum exposure during the pandemic? That meant we had to make some very tough decisions on continued investments in the products, which we are getting as a long tail in the portfolio. And at the end of it, even with the macroeconomic pressure due to the pandemic, we were successfully able to grow by double digits and almost double our operating margins, which was a huge success for us to secure future investment. That was the first part of it. The second aspect of the job was not just to harmonize the existing or transform the existing product portfolio but also to focus on launching new products in remote operations, in the space of connected workers, to help our customers fundamentally transform their way of doing the business. And at the same time, help Honeywell become- I would say move towards a higher-margin business of SaaS applications and ARR business. The strategy worked, and we saw some huge wins in a very short amount of time. If I must summarize this, these are the two fundamental areas where Honeywell has tremendous success. We are still seeing this even after I've left and joined Kelvin. They're continuing this particular path right now.

[00:17:56]

Ken: Nice story. And man, what an influential time. You're right about the pandemic. It's interesting; we often discuss our portfolio of companies. We've invested in all industrial IoT, but all of them fared well during that time. We came down to this pattern that everything we invest in is probably remote asset management. And that's the killer use case. And in some sense, you were forced to do that during that time because you couldn't send people out into factories and equipment. And so, in essence, across the industry, it was what we often call 'the great digital accelerator.'

[00:18:35]

Suhail: You're right. To call out on that aspect, we used to say that how can we allow our customers to run their factories from their homes? How can we allow them to run the factories from even a Starbucks? That was the whole intent of creating this digital thread and the technology that will enable our customers to get into this behavior when, as you mentioned, people cannot go into the factory. It's a very thin crew that can go inside the factory and manage the production. And that production itself is dynamically changing because the supply and demand needs are ever-evolving during the pandemic. It was great learning.

[00:19:11]

Ken: Yeah. Yeah, true. Remote access went from a nice-to-have to a must-have. And so, a lot changed very much during that time. And as we say, there's a new normal, and I doubt any of it's going back to the old patterns there. Speaking of new normal, you joined Kelvin as Chief Product Officer in May 2021. Tell us a bit about the company and your remit there.

[00:19:33]

Suhail: In my previous experience in launching products in the industrial IoT space, the first steps in helping the customer see the value of digital technology have been predominantly around solving the problems of IT and OT convergence problems. Getting access to the data, providing open-loop data and analytics recommendations to resolve issues, and even automating some of those workflows for the application users like, how can you automatically create maintenance tickets or digitize some of the SOPs or operating procedures for the field workers and help them while they're doing that particular job? For me, as part of the innovation, the natural next step for me was to take this knowledge and experience to move towards closing the loop in the controls plane, not just focus on the network and the data plane in an open loop, kind of recommendation and analytics only, but moving towards closing the loop with the controls plane to enable autonomous industrial operations. Think about it. If the company's working in the autonomous vehicle space, if they can automate the driver operations, can we take similar concepts and learnings to automate some of the complex operations in the manufacturing space? You may have noticed that I stressed the term collaborative control, a key aspect of the product. At Kelvin, we are focused on this aspect of the software stack. We provide our customers with a collaborative controls software platform that helps them automate their industrial operations by leveraging the concepts of industrial IoT, AI, machine learning, RPA, and Edge and Cloud convergence.

We are not trying to replace an engineer or an operator on the manufacturing floor or a production facility through advanced automation. The key to autonomous operations for us is first to have the ability to provide a single source of truth to the production engineers, the control engineers, to the operators on the health of the assets, the health of the process, and the systems for them to understand whether, within the given constraints, their production facility can meet that dynamically changing production and sustainability goals and KPIs. And where the bottlenecks are, what's causing an issue and how I can resolve them. Once these engineers have the single source of truth, this digital lineage, this digital thread- the software then allows them to communicate and collaborate in real-time to understand the root cause of the issue. Tag data and interesting events, perform root cause analysis, simulate complex optimization scenarios, and even automate some closed-loop control decisions to improve their operations. Every interaction with the software creates a training set that allows the software to become more intelligent over time and predict failures, providing recommendations to resolve the issues and even automating them- all without the need to upgrade their existing industrial automation stack or go through complex bespoke projects, which will lead nowhere across the organization. We are giving the power of modern AI-driven automation back to the hands of the engineers and operators and creating this assistive technology that will one day lead us to sustainable and autonomous manufacturing operations. And that is a pretty long answer, I know. But I hope that gives you a good perspective of why I'm so passionate about this space and why I joined Kelvin to build this innovative product and an ecosystem around that.

[00:23:12]

Ken: The answer is issued because it paints the vectors you brought in earlier about the Internet of Everything and the internet network access and how important that was and then clearly on the control side. And in some sense, it feels like you're merging all those. I know this will be very over-simplistic, but robotic process automation for OT makes a lot of sense because many process automation works up through traditional ERP systems in business processes. The best ones do. And looking at technology layers to help, in essence, bridge that but also automate that, especially using AI techniques or optimization techniques, makes a lot of sense. I'm surprised nobody else has been doing this before. Tell me a bit about some of the early use cases and wins your clients have seen using Kelvin.

[00:24:03]

Suhail: The good thing is that our work at Kelvin is not some moonshot vision a couple of decades into the future. This future of autonomous and sustainable manufacturing that we're talking about is happening right now. And the current strain the pandemic has put on our supply chain is only leading to the acceleration and adoption of this particular technology. I'll give you an example. We have a very forward-thinking customer in Australia operating thousands of oil and gas wells in the Cooper Basin. They are automating their production processes by optimizing the operations of various artificial lifts like beam pumps, plunger lifts, progressive cavity pumps, which has led to significant cost savings and production uplift. The more interesting aspect is by automating some of these processes, they've been able to reduce fugitive methane emissions by as much as 53 percent in some of those wells, which is huge when you look at the scale of the operations. We talk about millions of metrics of tons of fugitive methane emissions saved from entering the environment. And this is another aspect of our company and the product strategy. We are building this company, this software, to do good for the environment we are living in. We have this concept of this double bottom-line impact in the company that focuses on helping customers improve their production and productivity, which helps their bottom line. Still, at the same time, it helps them become more sustainable and meet their net-zero initiatives. Combining these two aspects helps our customers easily justify the ROI in partnering with us.

I'll give you another example. We have another customer who is one of the largest service providers in energy exploration, development, and production. They are currently using Kelvin software to optimize and automate many aspects of their fracking operations, from drilling to completion. They have standardized their complete North American fracking operations on Kelvin. They are running 50+ optimization and autonomous control application on our platform. This has led to the bottom-line impact of approximately 100 million dollars in the last couple of years through productivity gains and production in truck rolls. Now, this is a great success story of digital transformation; in my view, when done right, the customer sees tremendous value in this partnership. These are just a couple of examples out of many customer engagements we have globally, where we are working with some of the largest companies in the world in process manufacturing verticals like oil and gas, across upstream, midstream, downstream, chemicals, mining, food and beverages, CPG, in pharma- helping them optimize and automate their production workflows, were engineers working in these companies are using the tools provided by Kelvin and combining their expertise with the intelligence of the software.

[00:27:05]

Ken: I think you guys are one of the most impactful OT companies that most have probably never heard of. Kelvin was introduced to us probably about a year ago and started to hear more. People were telling me, you really ought to pay attention to this company. And, of course, we've gotten closer to it and have supported the company in several ways. But I must say, you're so quiet. I would think you're almost a Swiss company. You're maintaining a big presence but very quiet. Hopefully, your leadership on the product side will help bring more visibility to the great work and the great impact you guys are having on ESG goals. We mentioned several times in the conversation about the Edge and, particularly, the Industrial Edge. And as for your final question, it would be great to have you put your prognosticator hat on for a moment and give us a sense of what you think the next several years hold for what we call the Industrial Edge.

[00:28:04]

Suhail: We talked about that in the previous conversation, how the last few years have been more about Edge eating the Cloud. The good thing is that that argument is now settling down. One must realize that Edge and Cloud computing complement each other. It's especially true in the case of the operational data on a factory floor. It's a fact that in the next few years, over 90% of the operational data will be connected and processed at the Edge in a distributed environment connected via ethernet, Wi-Fi, or a 5G network. And this is going to create new technological paradigms. But if I have to summarize, I believe at least three areas I would highlight where we will see innovation in the next three to five years. The first is the innovation in distributed computing and machine learning at scale at the Edge. Beyond just provisioning and managing the Edge devices, I'm talking about the ability to connect, process, filter, and analyze real-time data streams in an unsupervised fashion at scale. This data could be the operational data coming in from the machine, so the sensors or even video sensors, which require complex computational analysis at the Edge, will see more and more companies follow this route where we are doing the unsupervised learning, especially with the new connectivity paradigms coming in, new compute paradigms coming in around CPUs and GPUs becoming more proficient. The hardware itself is becoming cheaper and more powerful to be deployed at the Edge. The second aspect for me is cybersecurity, which is essential. The technology stack is evolving- we are moving towards the concept of interconnected cyber-physical systems. The current security solutions must be re-architected to support a distributed cyber-physical architecture. This is another area where we will see more and more Industrial Edge devices and Edge computing play a role in cybersecurity- securing the data, the assets, the information from the bad actors. The third area that I would highlight is more around the enterprise and the operational workflows. They have been sitting there in the Cloud for a long time. And with all the new changes and the compute paradigms shifting on the Edge side, we will see more enterprise and operational workflows move towards the Edge.

I'll give an example. Suppose you can process and analyze data at the Edge. Can you allow the Edge systems to automate decision-making and close the loop? Think of RPA or Robotic Process Automation happening at the Edge. Another example is around the workflows related to augmented reality or AR on the factory floor. Now that the 5G technology can mitigate latency issues, they can think about integrating workflows related to connected workers at the Edge. A field worker should be able to receive contextual enterprise and operational data while going through the operating procedures like doing the inspection, doing maintenance, or doing some other job in the field. So more and more enterprise and operational workflows are going to move towards the Edge as the computing paradigm shifts. In my experience, I would say three highlights that will see how the Edge will evolve for the industry customers.

[00:31:27]

Ken: Very insightful. Today, we had a deep dive for one of our clients on Edge computing features. And you've not only highlighted some of the aspects we discussed but also provided some deeper insights on aspects of this, especially around this idea of RPA at the Edge, as you call it. In closing, I always like to ask a final question: where do you find your inspiration?

[00:31:50]

Suhail: The inspiration for me comes in many ways, to be honest. I believe strongly in learning from the doers in technology and digital transformation. I like to listen, read, and learn from the people who have done this in some aspect of it. Maybe if it is in a different vertical or industry, some interesting work is happening in the healthcare or financial sector in the industrial space. What has been done where the industry is headed? Can we bring some of the learnings from the enterprise and the consumer world into the industrial world? So, I would say I read a lot. Not just the books, but the articles and even some of the podcasts, including this Digital Thread Momenta podcast that we talked about right now. It's impressive to learn from the people working in the same or similar challenging environment and leading the transformation for those companies. That is my inspiration all the time. And you know what? When I was looking for another source of inspiration, I purchased a 70-year-old Ford Mustang, which has taught me a great deal of patience in repairing it, only to have it break down again two weeks later. If I want to practice patience, I go fix my Mustang.

[00:33:17]

Ken: I grew up with Fords myself. I think that you've found- on the road dead, fix or repair daily. You name it; I'll apply. But I tell you, I loved working on some of the older Mustangs, so it's down to the metal, as they like to say, in terms of simplicity. I can completely relate to you both on observing patterns in terms of the reading. I also will call it very traditional architecture in terms of Ford Mustangs. Look, Suhail, thank you for sharing this time and these insights with us today.

[00:33:49]

Suhail: Absolutely, my pleasure. Thank you for having me, and I look forward to hearing how this podcast comes out.

[00:33:55]

Ken: Well, actually, I believe from what I have heard up to this point, it will be one of our best to listen to. This has been Suhail Jiwani, Chief Product Officer at Kelvin AI, a company helping industrial customers quickly scale control applications that create sustainable business value. Thank you for listening, and please join us next week for the next episode of our Digital Thread podcast series. Thank you, and have a great 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 and resources to help with your digital industry journey. Thank you for listening.

[The End]