Why the Fastest Megawatt Is the One You Don't Use: Encycle's 21-Year Data Advantage | Ana-Paula Issa

Russ Fordyce (00:00): Welcome back to the winner's circle, everybody. I am here today with AP from Encycle. They are out there at the forefront of energy. And we're not talking about the supply side. We're talking about something else. AP, welcome to the show and congratulations on the Sustainability Award. AP Issa (00:17): Thanks, Russ. Nora, happy to be here and proud of the company for sure for all the work that it took to get there. Russ Fordyce (00:23): Yeah, and it has been a ride. So before we get into the guts of of the company and the nomination and everything you guys are doing, tell us just give us an elevator pitch on N Cycle. AP Issa (00:35): Ncycle an optimization company. we started out as an optimization company for energy efficiency of RTUs, so HVAC rooftop units. but we have grown a lot. So we now do demand response, automated demand response. So we're affecting both your peak as well as being able to respond to utility events, which obviously counts towards grid stability, which is awesome. and now we're in the process of working very closely with customers and optimizing workflows. using AI actually to optimize workflows through the facility from fault detection, literally cradle to grave, fault detection of units, getting them fixed and then tracking whether it was actually fixed and should I pay the invoice. so it's a it's a solution. And as a company, we end up working with both the energy teams as well as the facilities teams. Historically, those can be kind of siloed, but in reality, they actually they actually affect each other really significantly. and so yeah, so we are a software AI optimization company. That's what we specialize in. And where we see gaps in the marketplace and things to help people with, we build software for it. Russ Fordyce (01:53): Well and and you've been at the company a long time. You've been you started out in the sales department and now have risen through the ranks to grab that CEO title. And it's a it's a it's gotta be A humbling to do that, but B you've seen a tremendous amount of change in the last decade or more in this sector as everything comes aboard like AI. AP Issa (02:01): Yep. Yeah. AP Issa (02:12): I have AP Issa (02:17): Mm-hmm. No, I definitely have. when I first started with the company back in 2008, we we were very much a controls platform. not because we wanted to be. We already were playing with machine learning and AI back then. So everybody looks at AI as like the new thing, the new kid on the block. But no, some of us have actually been doing it for a while now. but at the time, you really couldn't connect to a building automation system. Not not at the level that you need to in order to actually run machine learning and AI algorithms in the background. You it just wasn't fast enough. Like we're dealing with if you go back to 2008, I mean you're dealing with building automation systems that require dial up. Like it's old. and so we were installing controllers on rooftop units. And I will say that those controllers gave us the amazing database and the depth of data that we have today, which is why we're so good at what we do today. but it took us a it kind of took us a while, like we were kind of ahead of the curve, and then we had to wait for the building automation world to kind of catch up with us. and now that we're into a position where building automation systems can actually connect to the cloud and you can get as close to real time as possible. Like we communicate every five minutes now and decisions are processed every five minutes. That that transition has really only happened in the last, I'd say probably five years since we're really starting to see it on the customer side infrastructure-wise. in Cycle, we pride ourselves on the fact that we don't we don't go and install hardware or anything like that. We are purely a software deployment. but that decision making protocol ha does allow for us to b essentially affect base load. so from a utility perspective. Russ Fordyce (03:51): That's the five or AP Issa (04:16): Like when you're looking at what's happening with utilities, so they're expecting like fifty-five percent increase in grid requirements within the next like decade, right? you've got data centers, you've got EV. EV's gonna grow at least two two thousand percent, they're estimating. Data centers gonna add another three hundred percent in terms of growth and growth requirements. you've got utilities that are struggling from a capacity perspective. They you really need as much technology out there as possible that can use existing infrastructure to reduce that base load. And so that's where we it's kind of where we play and where we've played for a long time. which is one of the reasons that we won this award is 'cause our sustainability numbers are actually they're they're solid. Russ Fordyce (05:03): Well that that's really kind of where I was getting at is this this industry, you know, I I I would grew up in telecom and tech, kinda alongside you. It's they're very, very similar. and they're facing they're they're facing a lot of the same constraints, which is E V's coming and it's gonna steal all of our electricity if the data centers don't get it first. But walk walk us through a typical deployment of of a customer comes to you, give us a sense of who that customer is and kinda what you come in and do. It's AP Issa (05:21): Yeah. Russ Fordyce (05:30): You know, I i i I liken it to kind of that those telecom days in the old days when you could come in and tear apart somebody's infrastructure and bill and save ten, fifteen, twenty percent just 'cause of their dropping stuff on the floor. AP Issa (05:40): Yeah. Yep. I mean I definitely did see that when I was on the IT side of the house. but typical deployment for us is actually it's really easy. Like back in the day when we were installing hardware, like that's that's painful and very grateful we don't do it anymore. but in terms of a typical deployment today, it it's really an either an API connection or a file load onto the EMS. It's simple, straightforward. We can literally bring up hundreds of buildings a minute, is is what we're down to. and because because of the fact that we're connected to so many buildings and we've connected to so many RTUs over the years, and that data set is so deep, we're actually able to start learning and deploying sustainability measures from day one. So Like when you talk to a lot of people that are doing like AR optimization, what have you, like there's a learning period. And for some it can be quite long. with us, it literally learns as it goes and it's comparing comparing back to trend data and normalizing against that. So it's actually quite sophisticated. It's one file, everybody gets the same one, so it's not complicated. We don't need to like try different things depending on different building types. Again, that's because we've been doing it for as long as we have and we're installed as in many places as we are. So that's been a real driver for us, but a real differentiator within the marketplace. 'cause we're as a result of all of our experience, like we're a lot more mature. we're kind of past the science project piece of it. Good. Russ Fordyce (07:25): Well, i it it's interesting you say that 'cause I'm what I'm finding is companies are are being reinvigorated by their legacy legacy data. That data now is worth all sorts of value. I don't I don't know what it's not dollars necessarily, but to you and to your customers, that stuff is awesome, right? Because now you know exactly what a twelve hundred unit building looks like 'cause you have forty years of data on it. So I imagine AP Issa (07:40): That's it. AP Issa (07:53): Yeah. Russ Fordyce (07:54): How did that play a role in your development of of the AI to kind of look at this data? AP Issa (08:00): So when we look, so this is the thing with AI, right? There's a lot of companies that are joining the AI game now or have within the last couple years. And that's great. Like we definitely all of that all of that invigoration within the marketplace is good. But the one thing a lot of people forget is day AI is only as good as the data set it's built on. Right? The model that it's trained on. Like it's not It it I kinda sometimes I kind of view it a little bit. I've got I've got young kids. It's kind of like a two-year-old. You've got a teacher to say thank you and please and you know, what are the boundaries? No, you don't leave the deep freeze door open because you want, you know, ten glasses of ice. Like it's it's stuff like that. AI is not that dissimilar, so it's as good as what you're gonna put into it is going to determine what you get out of it. Russ Fordyce (08:34): Same. Same AP Issa (08:56): Because we started like our first installs our the very first installs were in 2005. So we're at twenty-one years. Then we we got a real uptick between two thousand and eight to two thousand and seventeen with demand response, auto DR, everything that was happening in California as it rel related to Lawrence Berkeley and the open ADR standards. All of those installs were on actual controllers that actually submeted every single load that it was connected to. So we've probably got the most, the most in-depth database on any just about any type of make and model number of RTU across all the different weather zones, across all types of different buildings that you could imagine, from school districts to entertainments and restaurants to movie theaters to big box retail. Like we got it. our AI is built on that. That is the determining factor. all of the load that we virtual submeter today, we're within one percent of a utility grade submeter. and that that's hard to get. Like that is not an easy feat. And so, but it's all because the data is solid. we didn't build the AI And then go and try and collect two years of data. We'd been kind of playing with ulmarithmic control for a while while we were collecting data. And now we have this massive data set, which is really what the software that was launched in 2017, that is all built on that. it's built on that work. So it's at the end of the day, it's no difference to Salesforce. Anybody that's in sales, I've came up in sales. Anybody that's been in sales, is in sales. y everyone's interacted with Salesforce and you know, crap in equals crap out is kind of the same th AI's no different. Russ Fordyce (10:59): Well, and I'm wondering what does it look like from the customer side? I know you're connecting to these endpoints, but I imagine that customer data set's gotta be atrocious most of the time. I mean, I I can't imagine management companies spending bazillions keeping that data clean. AP Issa (11:16): They're they're getting better at it. So with us, we're when we do deployments and what have you, we're basically we take customers like RTU data, we take their energy data, we're feeding it into our models, we're collecting where the data in real time. So all of that gets fed in. we collect the data, we keep the data, we normalize the data, and we don't ever throw it out. Customers can access it anytime, but we don't we don't throw it out. so When we're looking at performance, we measure the building against itself and against its peers. so it's it's constantly looking at that piece of it. it's interesting. There was a time where if you ask people for things like can I get an energy report or can I get your utility rates or can I get your HVAC data, they didn't necessarily have it. But I have noticed a pretty significant difference over the last five years. Like there's a lot more emphasis on tagging, the requirements of tagging, kind of why you want to keep this stuff clean. And I think a lot of that's been driven by the fact that you do kind of have this intersection of technologies, right? You've got the building automation systems that you now can communicate with in real time. They're still not great at holding data, but as long as you can get that into a data lake, you can keep your data. but that's only as useful as how well it's tagged within that system. And then when you go all the way through to facilities where you've now got facilities people that are running preventative maintenance programs and they're trying to those poor guys they actually spend a lot of time do playing whack-a-mole because it's still a very reactive industry. But for them being able to tag stuff better so that it when they're looking at the ticketing system, it matches up with the EMS system. Which matches up to what the store managers or what have you are seeing. That is getting a lot cleaner because you do. Like you've actually got all of these APIs that are now all talking to each other, and they only communicate as well as the tagging and the organization behind. Our customers are really lucky. We we tend to normalize all of that for them. but they're they're getting better at it. We're we're seeing a shift. Russ Fordyce (13:36): Yeah, it it's interesting because the that same philosophy I think you've seen in other industries, manufacturing among them. And manufacturing spent a lot of effort and time over the last kind of two, two decades, three decades on automation. And I think a lot of those controllers and and sensors are in a lot of that manufacturing now to where those AI agents are actually going and doing that maintenance work because they're they know that that AP Issa (14:01): They are. Russ Fordyce (14:02): That engine's been running for 49,000 hours, right? And they're like, okay, at fifty, that thing blows or it needs maintenance or whatever. And so they're actually using that as kind of to get to that proactive model. And you think that this industry will get there as well pretty soon? I mean, I I can't imagine these large scale manufacturers are sitting by idle, not not wanting to give you this data. AP Issa (14:23): No, yeah. No, there's definitely there's definitely been a shift in the industry. So what I can tell you from what I'm seeing is that most of this industry when it comes to an HVAC unit, everything is very reactive. I think because of where it where it's located on a building. HVAC units on the rooftop of like commercial manufacturing schools, they're not visible. So unless it's smoking, leaking, on fire, or someone's actually uncomfortable, no one's doing anything with it. Like it it's so the triggering event is usually pretty expensive on the repair side. and so there's kind of Russ Fordyce (15:10): And that is true that's so true on the residential side too, having put in new HVAC systems. This would really be not the hardest thing in the world to solve. AP Issa (15:15): It's the same. One interesting thing. AP Issa (15:21): No. And so that is like as part of the award that we received was actually for chorus and chorus AI as well as for our optimization piece on the sustainability side. Because we're able to use trend data and we've been keeping trend data, I mean, since 2008, our trend data, that trend data has allowed us to make informed discussion, informed decisions on our to use. To the point that we can actually detect a unit in the process of failing, like 45 days before it actually fails. So before it starts smoking or leaking or goes on fire or whatever the other triggering event was. well, because the whole time it's failing, it does use more energy. And it does get to the point before it completely fails where it's using energy and literally doing nothing with it. Like you're spending money to just dump hot air into your building. Russ Fordyce (16:03): Because it's sucking energy. AP Issa (16:20): So it's it's useless. so all of our trend data really does inform that. And getting ahead, like if you were to think, right, if you're a school or you're a big box or you're a movie theater, you're all finding out that you got failed units at around the same time if you're using the average model today. Which means you're all competing for a mechanic to come on site and fix your unit. You're all competing for, yeah, it's very expensive. You're all competing for parts. Like it like it becomes a bit of like the snowball effect. So if you'd known about the problem in April versus knowing about the problem around Memorial Day, that gets you ahead of all of that. So that means your your reactive maintenance on your PL comes down. Like that that cost line comes down significantly. You get parts ordered ahead of everybody else. So your buildings are actually cool during summer. because RTUs themselves are kind of th they're dumb. They they they have two pieces of information, which is this is where I need to be, this is where I am right now, and it closes and opens relays based on that. Like it's there is nothing sophisticated in it. It certainly can't tell you that it's broken. Like it'll still close the relay, it will still run. Which is still going to cost you energy. It adds to the grid. It does not help with sustainability because you're using more than you should. so if you can get ahead, and that's why I say facilities teams and energy teams, even though they don't necessarily report up the same channel, they don't always work closely together, they do affect each other's jobs significantly. Russ Fordyce (18:12): Yeah, it reminds me of when we had IT and telecom. You know. W one was the data closet and one was the phone closet, and they were two different closets at one point, folks. AP Issa (18:17): yeah. AP Issa (18:23): They were. They're not anymore. And they weren't even back when I left the industry in two thousand eight are already kind of come all the way together. But yeah. It did it. Russ Fordyce (18:25): No. Russ Fordyce (18:31): Two thousand kind of closed that out. What when you think about when you think about these large scale deployments, you're t you're talking about how you can bring hundreds of buildings up, bring these large portfolios up on your software. What's the big thing that can that your customers kind of get surprised by in that first, you know, seven day, thirty day period? AP Issa (18:41): Mm. AP Issa (18:52): They usually get surprised by the amount of data flowing in. It's a like the data we get is really, really granular. and they're really surprised by the level of connection. So, like we can see things that you won't even see in an EMS. EMS is really driven by this is what my schedule should be. This is what my pre gr programming is supp supposed to be doing. But it doesn't tell you much more than that. And then you get a list of alarms that all fire to you a thousand times a day and nobody looks at. when they start to see the way that we kind of pull the data in and then correlate the data for them, and they get to see things like, for instance, my EMS has a schedule, it's running the schedule, none of my RTUs are listening to it. They're just on, regardless of time of day. That's a real eye opener because a lot of people don't you don't see that. You just assume, well, if I programmed my schedule a certain way, it's going to run that way. There's a lot of that, those aha and gotcha moments. And I think also just from a trend perspective, a lot of a lot of customers are kind of they're stuck in this preventative maintenance model where they check, you know, they check bands and filters and all the rest every couple of months. And they're kind of hoping for the best. And then they're reliant on hot calls. And I think what we've seen with a lot of customers that have actually started to be to use the trend data to drive decision versus the hot call is that they're seeing a forty-four percent reduction in their hot calls going in summer, which is huge. And that's like right out the box. That's that's before you start to look at what does year two, three look like. when you've really got into this thing about getting ahead of the real problem. but it's the level of data and the amount of information that is there and the way it's correlated in such a way that actually does drive a decision. In energy, much like in IT, when we were in it, we don't have it really a data problem anymore. Everybody's got data. AP Issa (21:14): We've been collecting data for a long time. There are a lot of companies that had a data lake for at least the last 10 years. Data's not the problem. It's what's the decision that the data drives? And that that has not been solved for by a lot of folks. They're still they still have all of this data and still asking, well, what now? So that that's really where we're starting to like see a lot of our growth with our customers is like the what now because we can answer that question. We can drive the decision based on the data. Russ Fordyce (21:49): Well, and that and that's what you want is you want a system or a consultant or you know, a service that's gonna help you mature, not just fix kind of the immediate problems. What are you seeing in terms of savings that year one? I mean, this is we are we're not in an energy crisis, but boy, does it feel like we're going into one. I mean, people really don't under AP Issa (22:09): Yeah, we're we're yeah. Russ Fordyce (22:11): It it it is the AI is eating us up and if you look at companies like Generac and other companies that were struggling before because they offered a different type of energy, they're not struggling anymore, right? These guys are s these companies are looking for energy any way they can get it. AP Issa (22:21): Mm-hmm. No. AP Issa (22:28): There's it it's not the like it's not the energy crisis that we saw like off to Enron, right? Like it's it's not quite that. But we are facing an energy crisis. It is not it's not as cut and dry, should we say, as it used to be. we do not have capacity on the grid. We just don't. And when you look at all of the changes that are coming down from a legislative standpoint, from you know, like like here in California, we're being legislated that we have to have EV. Our grids are necessary, like, right. And if you if you start to look across the United States where you're having those mandates, but the grid isn't ready for it. In California, we're really lucky our solar is like solid all day long, but our overnight load is maybe not the cleanest. And that does tend to be when people charge their cars. Russ Fordyce (23:08): Yeah. AP Issa (23:26): So, but we're kind of seeing where legislation is maybe getting ahead of where the grid is and the utilities don't tend to move quick enough to keep up with it, which is and it's interesting because like in the last like five years ago, even ten years ago, like my first some of my first meetings in California when I moved out here were with utilities talking about the EV crisis that was coming. So like everybody was watching the EV problem and they were like, this this could be bad. Nobody saw the data center train. Like like that was that came. Yeah. So nobody saw that piece. And that what that has done is really taken a problem that they knew that they had to deal with. And they knew the utilities knew that they were going to be behind already. And it's compounded that problem. Russ Fordyce (24:04): Yeah, welcome. It was right in their backyard too. AP Issa (24:26): significantly, which is why we're seeing the capacity. So like if you look at PJM, which everything underneath New York, basically down to the Carolinas and then across to Illis, it's like a like a pizza slice. That whole area, it used to be great if you were doing demand response because you would enroll in those programs and they just wouldn't call it. They'd call one test event and that was it. And you would be paid your DR participation based on that one test event. Russ Fordyce (24:36): One company. AP Issa (24:55): This year we have seen those programs called repeatedly because they do not have the capacity. and getting a grid up and running again after it's gone down, that's expensive and it's hard. and the thing is, is if if the utilities can't solve that capacity problem, then they're going to force it on customers. Like if you're not actively trying to reduce your overall load, if you're not actively participating in demand response. They're just gonna penalize you with capacity charges and those are gonna hurt. I actually think if you look at the US over the next five years, I th I think it's gonna get really uncomfortable really quickly. And I think a lot of people are gonna see the impact on their P and L. Russ Fordyce (25:38): Yeah, and if energy costs couldn't go higher. Yeah. yeah. AP Issa (25:43): Well, they've already gone up. So bet I think between twenty ten and twenty nineteen, we only saw like a ten percent increase. Like over that whole period, only ten percent. Between twenty twenty to the end of last year, we've seen a twenty nine percent increase. And just in the first six months of twenty six twenty twenty six, the utilities have already submitted eighteen point six billion dollars in increased energy rates that they're planning on in s on pulling the trigger on going into twenty twenty seven. So it doesn't I don't I don't know if everybody realizes that we're in an energy crisis, but one way or another we're all about to feel it. Russ Fordyce (26:23): Well, and and we should all thank each other for not buying as many EVs as we did. because it would make it a million times worse. And the I I d you know, I've heard a lot of stories from EV manufacturers. You know, it's the when those when the EV trucks come online, our cities are doomed. Those trucks consume more power than cities do. And so recharging those is gonna be a really, really big challenge. It's great for everybody, it's great for AP Issa (26:42): Yes. AP Issa (26:47): They do. Yeah. Russ Fordyce (26:52): You know, the environment. But boy, consumers are gonna really it's gonna be a trick. I I do love that legislation. You've seen it a couple of times of especially with data centers though, that they have to have onside power. And that's that seems like the only way consumers are gonna get protected. AP Issa (27:07): I think it is the only way, and a lot of that's been driven by the fact that utilities just can't move quick enough. The utilities are slow, right? They're just they just are. They always have been. I you know, do that would it be nice if there could be an easier way to solution with them? Yes. but I you know, this is this is definitely going to become a compounding problem as we see more and more load. And also the electrification, when you talk about if you go look at the different states, I mean. There's lot of states that don't allow gas heating anymore, that it's gonna now be electric heating. All you're doing is adding to a problem before we've even really found the correct solution. Russ Fordyce (27:49): Yeah. Well A P, thank you so much for all your expertise. This has really been fascinating learning about all the the great things going on in the grid and I'm a little scareder now, so that's probably that just means there's good opportunity out there too. I mean, you know, th it's great opportunity for folks in the energy business. It's never been better. AP Issa (28:00): Sorry. There's there is good opportunity. Yeah, no, there's good opportunity. And you know, I think if I think if we take very much a in it together approach where we're all kind of doing our part, I think that make like that makes a big difference. if it all of us kind of come together, reduce the amount that we're using, just be a little bit you know, people used to talk about, well, what is my carbon footprint? I mean, if you kind of take that approach and you're like, okay, well. Do I really need to have all these lights on? As a residential customer, do I really need my house to be at 70 degrees? Probably not. You you could probably do a little bit higher and be fine. from a commercial customer standpoint, I mean, they're g there's just gonna be this force where you're gonna have to go into real-time pricing. You're going to be forced to do demand response events. But if you can find technologies that don't require an expensive rip and replace that you can just put technology in. That will work with current infrastructure and reduce overall base load, like that is a huge win for the utilities and for consumers. So, like there's stuff, there's definitely opportunity out there. And I will say, whenever there's like these times of constraints, is always when we see the best technology is like really being embraced. So it's going to be interesting. I am it it's an exciting time to be in the market. Russ Fordyce (29:28): really is. Well, congratulations again on the win and we'll keep an eye on you going forward. AP Issa (29:32): Thanks. Yeah, no, I appreciate it. Thank you for the time.