Ryan Chacon is joined by SymphonyAI CTO Vijay Raghavendra on this episode of the IoT For All Podcast to debate AI’s function within the IoT trade. Vijay begins by introducing himself and the corporate earlier than speaking in regards to the significance of AI in IoT. He then talks extra particularly about AI’s match into enterprise know-how and learn how to begin adopting it. Vijay and Ryan then go extra high-level with conversations round challenges to adoption and the way value impacts the trade’s development earlier than ending the podcast with Vijay letting us know what to search for from SymphonyAI sooner or later.

About Vijay

Vijay Raghavendra is a know-how chief and entrepreneur with in depth expertise main know-how groups in firms starting from startups to Fortune 1. Most just lately, Vijay was the CTO at Acuity Manufacturers, an industrial know-how firm. At Acuity, he was answerable for all facets of software program know-how technique and supply, together with edge computing and IoT. Earlier than that, Vijay served as SVP of service provider know-how at Walmart, the place he and his groups had been answerable for all platforms, purposes, and algorithms that drove the expertise for Walmart’s retailers and suppliers and a big a part of the shopper expertise throughout shops and on-line.

All in favour of connecting with Vijay? Attain out on Linkedin!

About SymphonyAI

SymphonyAI is constructing the main enterprise AI firm for digital transformation throughout crucial and resilient development industries, together with retail, shopper packaged items, monetary companies, manufacturing, media, and IT service administration. SymphonyAI companies have many main enterprises as purchasers in every of those industries. Since its founding in 2017, SymphonyAI has grown quickly, approaching 2,000 proficient leaders, knowledge scientists, and different professionals. SymphonyAI is a SAIGroup firm backed by a $1 billion dedication from a profitable entrepreneur and philanthropist, Dr. Romesh Wadhwani.

Key Questions and Matters from this Episode:

(01:54) Introduction to Vijay and SymphonyAI

(06:23) Position of AI in IoT

(08:57) How AI matches into enterprise know-how

(14:00) Easy methods to begin adopting AI

(18:27) Challenges to adoption

(22:38) How does value have an effect on adoption

(26:42) What to look out for from SymphonyAI?


– [Voice Over] You might be listening to the IoT For All Media Community.

– [Ryan] Good day, everybody, and welcome to a different episode of the, IoT For Al Podcast, the primary publication and useful resource for the Web of Issues. I’m your host, Ryan Chacon. I do ask in case you are watching this on YouTube to please give this video a like, and subscribe to our channel, for those who haven’t already achieved so. And for those who’re listening to this on a podcast listing, please be at liberty to subscribe, to get the newest episode as quickly as they’re out. All proper, on as we speak’s episode, we now have a Vijay Raghavendra, the Chief Expertise Officer at SymphonyAI. They’re an organization that’s constructing a number one EnterpriseAI firm for digital transformation throughout crucial and resilient development industries. Together with retail, shopper package deal items, and lots of others. Actually good dialog right here. We discuss so much about sort of AI, the function of AI in IoT. We discuss in regards to the applied sciences that they’ve constructed, the applied sciences that they work together with frequently, how folks can get began with adopting AI into their resolution. Why vertical particular experience is admittedly essential, with regards to that sort of integration and bringing AI into an answer, is especially an IoT resolution. And we additionally discuss so much in regards to the challenges that they see from their aspect of issues, because it pertains to bringing options to life. So, all in all nice dialog, a whole lot of worth right here. I believe you’ll get pleasure from it. However, earlier than we get into it Works With, by Silicon Labs has emerged because the go-to developer convention for constructing the talents wanted to create impactful linked units. On September thirteenth via the fifteenth, Silicon Labs is bringing collectively influential, know-how manufacturers, ecosystem companions, and builders for 3 days of technical coaching and workshops, keynotes and professional panels. Works With is dwell on-line and free. Register at workswith.silabs.com workswith.silabs.com And with out additional ado, please get pleasure from this episode of the, IoT For All Podcast. Welcome Vijay to the, IoT For All Podcast. Thanks for being right here this week.

– [Vijay] Thanks so much, Ryan. Nice to be right here.

– [Ryan] Completely. Very enthusiastic about this dialog. I wished to kick it off by having you give a fast introduction about your self, to our viewers, for those who wouldn’t thoughts.

– [Vijay] Nice. I’m Vijay Raghavendra, I’m the CTO at SymphonyAI. I got here to SymphonyAI about seven months in the past because the CTO and previous to Symphony, I used to be the CTO at an industrial know-how firm referred to as Acuity Manufacturers. Acuity, along with being one of many largest industrial tech participant with the lighting and lighting controls additionally has a linked units play for constructing administration, for location administration methods as effectively. And I spent a while working with Acuity to construct these capabilities out. And previous to Acuity, I spent about seven years at Walmart, a number one varied components of engineering and product at Walmart. And I got here to Walmart as via an acquisition of an organization the place I used to be a co-founder and CTO that I offered my co-founders and I offered to Walmart.

– [Ryan] Unbelievable. Yeah. Very in depth background expertise. Seems like a fairly enjoyable journey to get to the place you are actually.

– [Vijay] It’s been, yeah.

– [Ryan] So let, let me ask you this. So let’s discuss SymphonyAI actual fast. Inform our viewers a little bit bit in regards to the firm, what the main target is, the function you all play in IoT, that sort of factor?

– [Vijay] Yeah, so SymphonyAI is an EnterpriseAI firm and our focus is to use AI and machine studying to resolve issues in varied verticals that we play in. From retail to monetary crime, to industrial to media, IT companies and federal. So, the main target for us is to allow our prospects in every of those verticals to actually rework what their companies and resolve actual issues via the appliance of AI and machine studying. The entire work that we do is grounded in our AI platform that we name Eureka. We not solely assist all the capabilities that you’d count on from any of the AI platforms, however we even have some distinctive capabilities round know-how that we’ve constructed particular algorithms, equivalent to topological knowledge evaluation. And particularly for every of those verticals. One among our crucial methods wherein we add worth, or we convey worth to our prospects is thru the deep vertical experience and our pre-trained fashions and capabilities that we now have throughout these totally different verticals. And particularly coming to the appliance of IoT, relying upon the vertical take industrial, for example, we work with some very giant manufacturing corporations and browse knowledge from a lot of totally different sensors. As you may think, in a whole lot of manufacturing facility to then use the info, to determine, to foretell outcomes equivalent to when a compressor could also be going dangerous. How do you then do preventive upkeep on these gear to do it, to forestall a a lot greater drawback from taking place downstream? And with that, we usher in all the capabilities, not simply with all kinds of various sensors and units that we learn knowledge from, however edge computing, digital twins, and deep studying fashions within the cloud.

– [Ryan] Completely. Yeah, unbelievable overview. Thanks a lot for sort of giving us a little bit little bit of context there. I do wanna ask you although, simply from a excessive stage standpoint, once we discuss AI and IoT, they oftentimes are actually going extra hand-in-hand than ever earlier than. Inform me in regards to the function and the way you view it of an AI firm within the IoT house?

– [Vijay] Yeah, I believe that’s an amazing query. And I believe for the longest time, I actually consider that folks have at all times believed within the worth of IoT and the potential of IoT. And you’ll see that within the shopper house. And I believe that possibility of IoT within the shopper house has, clearly, it’s turn into mainstream now, however within the enterprise house, we’ve at all times been on the cusp of realizing the worth. And up till just lately, I’d argue that we hadn’t totally realized the worth, however with all the modifications which can be taking place and have occurred, with the price of {hardware} coming down with the density of the compute in edge units. Getting to some extent the place it turns into actually fascinating for us to do a whole lot of processing on the edge with the power now to embed a TPU, for instance, in an edge system. So we are able to run tiny ML fashions on the edge, mix it with operating a deep studying fashions or extra subtle fashions within the cloud after which pushing the outcomes again. I actually suppose we now have all the underlying capabilities we have to actually convey the facility of AI mixed with IoT to resolve actually fascinating issues, such because the one I discussed just a bit bit earlier. So, I actually consider that AI turns into the conduit for unlocking the worth from IoT, as a result of with out the power for us to do one thing fascinating and helpful with all the knowledge that you simply’re getting from these units, it turns into clean. And with all the modifications within the {hardware} with cloud, with edge, with the development and AI, I actually suppose it’s a good way for us to convey the facility of IoT and resolve actual issues.

– [Ryan] 100% agree. So inform me a little bit bit about once we discuss AI, you recognize, there’s plenty of firms on the market who do AI, say they do AI, there’s plenty of options on the market. And oftentimes for the those who wish to convey AI into their resolution, they don’t at all times know precisely what the distinction is between firms that play within the house that the choices, what they need to be on the lookout for. So inform me, it’s mainly, it’s two questions I’ve. One is to inform me how, what you all do sort of differentiates from different gamers out there? And the opposite is how does this sort of know-how actually slot in to the way forward for enterprise know-how as a complete?

– [Vijay] Yeah, so, I believe you hit on a extremely crucial level with the appliance of AI and enterprises. I believe there are a selection of research that present that 80 plus % of all AI initiatives or AI initiatives in enterprises which can be attempting to undertake AI fail. And so they fail for a lot of totally different causes. Beginning with actually a lack of knowledge of those particular sort of issues there’s attempting to resolve for, for which AI and ML are a extremely good match, not having the appropriate knowledge or knowledge tradition, and never having the appropriate possibly mindset or modifications within the processes that they should do to actually reap the benefits of this know-how. So, I can maintain going. However the truth is that a whole lot of enterprises wrestle as we speak with the appliance of this actually fascinating half know-how to resolve issues. So, how can we, or the place can we play and the way can we assist our prospects actually get previous this crucial problem? As I discussed in my intro, one of many crucial locations the place we’ve invested as we’ve constructed out our merchandise, and product providing within the totally different verticals is the deep vertical experience that we’ve constructed over time. So, we aren’t a generic AI or ML platform with just a few fashions and we are able to throw it over the fence to our prospects and say, “Nice, go have at it.” What we’re targeted on due to our experience is with each single vertical, we’re tackling the precise set of issues the place AI and ML are an excellent match that. And the place we are able to go assist resolve these issues in a singular and differentiated method. So take one thing like determining which assortment you need to be carrying in a retail retailer and the way a lot of every assortment you need to be carrying and the place? That could be a very particular drawback that each single retailer has to resolve, massive, giant, or small. And what we’ve achieved with our deep experience and experiences, we’ve constructed over time these pre-trained fashions that resolve this drawback for retail and retailers and CPGs. And with the partnership that we now have with our prospects, we then begin from our pre-train fashions that aren’t simply the fashions and the options, but in addition embed the information of what a service provider at a big retailer does. How do they consider what the appropriate assortment is? What the combo is? We embed that intelligence to then work with our prospects, to then optimize these fashions and options to resolve the issues. And we do the identical factor with monetary crime and anti-money laundering, for instance, and so forth. And as in, so doing, we convey some very distinctive IP. I discussed the topological knowledge evaluation. Which is a extremely distinctive and fascinating method to consider knowledge and discover knowledge in an unsupervised studying method. Which takes very giant dimensional knowledge and permits us to suppose, have a look at this knowledge and discover relationships that won’t in any other case be apparent or that ordinary clustering algorithms could not provide you with. So, all of those collectively permits us to actually differentiate ourselves from others and deal with fixing the issues for our prospects.

– [Ryan] Yeah, it makes complete sense. One factor you talked about in there that I truly wished to observe up and ask you about is you had been speaking about sort of that vertical, particular experience of the area experience that’s tremendous invaluable and sort of the differentiator for you. I do know once we discuss to, let’s say platform firms within the IoT house, that’s an enormous factor for them to assist separate themselves out is versus simply having this common platform that may do all of it, or at the very least that’s how they promote it. They discovered worth in buying prospects with extra of a focused focus based mostly on area expertise that they’ve for fixing a specific drawback or explicit use instances inside an trade. So let me ask you when a listener to that is seeking to sort of be taught extra about getting AI components concerned of their resolution. They’re seeking to undertake AI know-how. How ought to they sort of go about getting began down that course of and why is it so essential to discover a firm with the vertical particular experience that connects to them to assist simply enhance the chance of success?

– – [Vijay] Yeah. Wonderful query, once more. My recommendation to firms that wish to incorporate AI to resolve issues could be for them to be very clear and spend the time to actually perceive that the outcomes they’re attempting to have an effect on and the precise sort of issues they’re attempting to resolve and actually get educated both via partnerships or working with firms, equivalent to ours to essentially perceive the kinds of issues for which AI is an effective match, as a result of it isn’t a panacea for each drawback that each firm’s gonna have. So, it’s actually essential for them to know that. The second is, and I can’t emphasize this sufficient, AI or any of those machine studying algorithms simply don’t work, or it means nothing for those who don’t have a tradition of excellent knowledge and a tradition of serious about knowledge as the important thing enabler for the appliance of AI. If it’s actually rubbish in and rubbish out. So for those who don’t have good clear knowledge, and in case you are probably not listening to that as a core elementary a part of the way you construct your merchandise and your methods, frankly, nothing else is gonna work.

– [Ryan] Proper. So, I’d say I’d actually encourage firms to actually essentially deeply perceive each the issue, but in addition then deal with the info and guaranteeing that they’re not solely have the appropriate knowledge, however they’ve a tradition of guaranteeing good clear knowledge as a result of that then turns into an enormous unlock. After which, clearly, specializing in not simply broad enabling platforms, which more and more have gotten a commodity, however working with and partnering with firms that may actually usher in that very particular area experience turns into a key approach to guaranteeing that you simply’re fixing issues that matter as a result of, finally, the constructing blocks for a way you do convey knowledge from varied sources, clear the info and rework the info and the way you construct the fashions themselves, are more and more a commodities. The secret’s gonna be, do you actually perceive the verticals and the domains? So you’re extracting the appropriate options. You perceive when knowledge or the fashions are drifting. All of these key issues are key concerns, I believe are what’s going to make it profitable.

– [Ryan] Completely. Yeah, completely. Let me ask you this slight little pivot right here to a little bit totally different space of focus, however while you sort of have a look at the evolution of the know-how, not simply in IoT, but in addition in AI and sort of how they work collectively, what do you suppose have been among the largest challenges to getting the know-how the place it must be? To extend adoption, to sort of, you recognize, these expectations or these numbers that we’ve been promised for therefore a few years in regards to the development and adoption of those applied sciences, what do you suppose have been the largest possibly roadblocks or velocity bumps which have sort of received in the best way to that sort of main as much as the place we are actually?

– [Vijay] Yep, I’d say, for those who suppose again to the final decade or so, as I discussed a short time again, I believe the promise whether or not it’s with IoT and even broadly with enterprises has at all times been there, however the adoption, and extra importantly, the success from the adoption of this know-how hasn’t fairly saved up with the expectations. And the explanations are, I believe, once more, a elementary, possibly lack of knowledge of how this know-how works. And as I discussed, the deal with knowledge, and I do know I’m harping on this, like fairly a bit however I’m doing it as a result of it’s so essential. I believe firms, on the whole, have underestimated the significance of getting an excellent knowledge, however extra importantly, having a tradition of excellent knowledge that’s inherent within the firm. And while you don’t have that, it turns into very tough to appreciate the worth actually at scale. So, I’d say that’s possibly one of many largest elementary challenges that I’ve seen. The second is I believe, as we’ve developed, particularly within the final 5 to seven years, the capabilities that we now have from whether or not it’s from cloud frameworks to different open supply frameworks, to a lot of Python libraries, to transformer fashions that are actually accessible, or has essentially modified the sport. To the purpose you don’t want a PhD in math and stats and laptop science to start out constructing and realizing worth. So, I believe the know-how has additionally developed actually quickly within the final a number of years, that makes it far more fascinating, far more tractable drawback to resolve. And let’s face it the final and the largest drawback is at all times gonna be for us to search out good expertise at scale and knowledge science and ML Expertise might be one of many hardest expertise for us to search out. And that’s the place the power for us to have the ability to leverage a whole lot of these open supply fashions for a citizen knowledge scientist, for instance, to have the ability to use a whole lot of these fashions and options to resolve enterprise issues while not having a staff of information scientists. I believe all of those collectively will assist unlock the worth quicker.

– [Ryan] Yeah. Completely agree. I imply, there’s new applied sciences day-after-day, proper? You realize, BLE Wideband, Extremely-Wideband. Edge computing’s turning into extra highly effective, the cloud. All very massive enablers of what we’re sort of speaking about. How do you suppose the price ingredient components into the sort of adoption? I imply, clearly value appear to be happening throughout the board for IoT elements tech. Whether or not connectivity, the {hardware}, or the software program, you title it. How do you suppose that mixed with the evolution and development of the know-how facet is enjoying a task and influencing the longer term development of what we’re attempting to construct?

– [Vijay] Yeah. I believe value, particularly with IoT within the enterprise is gonna be an enormous issue. Curiously, among the work that I did at my earlier firm, and that was simply having a dialog with one of many firms that was utilizing Extremely-Wideband to do look monitoring, for instance. The size for, for those who take retail.

– [Vijay] The size at which at the very least a big retailer who has a number of places, the size at which they function, it turns into the price of the {hardware} turns into a really, very materials value, particularly for firms which can be working on very small margins to start with. And I believe that’s the place it needs to be a mix of value of the {hardware} has to maintain coming down and we now have to far more environment friendly in regards to the density of the {hardware} or the beacons. And so forth that we want in very giant areas. The, as you talked about, edge computing and the density of the {hardware}, the facility of the {hardware} on the sting is turning into increasingly amenable for us to do a whole lot of processing on the edge, which then helps us with the egress and ingress cross to the cloud. However, finally, I believe we’re at some extent the place all of those are trending in the appropriate method. The fee is coming down. The know-how with Extremely-Wideband, BLE, VLC with the NextGen of what’s coming with 5G for the enterprise are all, I believe driving efficiencies and have gotten create enablers for purposes that we are able to resolve. And finally, if the purposes that we are able to construct on all of those applied sciences, doesn’t ship sufficient worth to justify the worth then clearly it’s gonna fail because it ought to. However my agency perception is we’re at some extent the place we now have all of those constructing blocks. The fee regardless of among the challenges, for instance, for a big retailer is at some extent the place it is rather a lot within the ballpark of being very manageable for a retailer. And the worth that they will get from it, I believe, is justifiable and can solely get higher any longer. After which you’ll be able to apply the identical analogy to varied verticals as effectively.

– [Ryan] Completely agree. Completely agree. Yeah. It’s all excellent factors. I imply, the expansion and the brand new applied sciences popping out is a large half. The fee happening is an enormous enabler. The simply developments throughout the board and all the things occurring. The extra use instances, the extra profitable deployments, the extra area experience that we talked about earlier. Like simply the extra accessible that information and data is, and applied sciences is for folks, the extra choices they should construct the appropriate resolution for his or her explicit use case. So, as we wrap up right here, I wished to ask sort of wanting ahead from the place we are actually popping out of SymphonyAI, and sort of what you all have occurring. What’s the massive subsequent step? Like what ought to we be on lookout for listening to popping out of your man aspect of issues?

– [Vijay] Yeah, so one of many locations that we’re targeted on is, as you consider the vertical particular experience and fixing issues in numerous verticals, particularly with the AI and at the side of the usage of IoT and different units, the best way we’re serious about the issues we’re fixing and what’s coming subsequent is admittedly the notion of an AI enabled digital employee who’s actually working in live performance in-hand-in hand with the people within the loop, if you’ll. And actually enabling the people who’re working, whether or not it’s a enterprise analyst, a enterprise professional, somebody on meeting line, or a retailer supervisor to do their jobs a lot, far more effectively than they will as we speak with what are possibly primary analytics and even some primary AI enabled purposes. In order that’s the place we consider we’re going to see the true mainstream, not simply adoption, however a step perform enhance within the worth that this know-how can actually drive in with the appliance. Which brings collectively IoT and all the ecosystem round it with edge and the cloud with AI and ML, with the vertical experience. Actually bringing all of these collectively in a method that basically augments and helps the people in a elementary method. Which makes it in a seamless elementary method, I believe is the unlock.

– [Ryan] Couldn’t agree extra. Yeah, it’s very thrilling stuff. What you all have occurring over on the firm may be very fascinating from the analysis that I’ve achieved. I hope our viewers takes a while to kinda look into what you will have occurring. For our viewers on the market who does have questions, could wish to observe up, be taught a bit extra, what’s one of the best ways that they will do this?

– [Vijay] So, our web site symphonyai.com is a good useful resource. There are hyperlinks there. There’s a whole lot of actually nice content material for the issues we’re fixing in every of the verticals. And there are hyperlinks there for anybody who desires to ping us with any query as effectively. So that may be an amazing place for any of the listeners to ping us.

– [Ryan] Unbelievable. Effectively, thanks a lot to your time. Actually admire it. We don’t have the chance to speak so much about AI recently. So I actually admire you taking the time to do this. I believe our viewers is getting a ton of worth out of this dialog. So thanks once more, and we hopefully like to have you ever again, different members of your staff again to proceed this dialogue and discuss extra about how AI and IoT actually work collectively to take issues ahead.

– [Vijay] Nice. I admire the chance, Ryan. It was nice to speak to you. Thanks.

– [Ryan] Thanks. All proper, everybody. Thanks once more for watching that episode of, IoT For All Podcast. In case you loved the episode, please click on the thumbs up button. Subscribe to our channel and you’ll want to hit the bell notifications so that you get the newest episodes as quickly because it turn into accessible. Aside from that, thanks once more for watching and we’ll see you subsequent time.


Subscribe Us to receive our latest news in your inbox!

We don’t spam! Read our privacy policy for more info.


Please enter your comment!
Please enter your name here