Decarbonizing the grid with virtual power plants: Recurve’s Matt Golden

Decarbonizing our energy grid will take a lot more than adopting alternative power sources. That’s the founding principle behind Recurve, a company on a mission to speed the transition to a carbon-free economy, while making the already fragile energy grid more reliable. Aaron meets with Recurve CEO Matt Golden to discuss why decarbonizing the grid calls for a new approach to flexible demand, through what Recurve calls “virtual power plants.”

The discussion highlights the importance of demand-side flexibility and load balancing, the complexities of modernizing a grid, and how Google Cloud is making Recurve's mission possible.

Golden is a recognized thought leader in the energy efficiency industry. Whether testifying before Congress or auditing Ellen DeGeneres’ house, Matt has been spreading the word about energy efficiency for years. An entrepreneur, policy advocate and efficiency agitator, Matt is committed to bringing demand flexibility in the built environment to scale as a grid and climate resource. He has made presentations on emerging demand flexibility markets to the US Senate, at COP 21 and to the UK Parliament.

Featuring: Matt Golden, Founder and CEO of Recurve
The Catalyst by Softchoice is the podcast shining a light on the human side of technology.

This episode is brought to you by Google Cloud.
Speak to Softchoice today to see how we can help you use Google cloud to break down silos, unlock new insights and power innovation.

Transcript

 [00:00:00] Aaron: This episode is brought to you by Google Cloud. Speak to Softchoice today to see how we can help you use Google Cloud to break down silos, unlock new insights and power innovation for your company. Welcome back to another episode of The Catalyst by Softchoice, the podcast dedicated to exploring the intersection of people and technology.

[00:00:20] Aaron: And I'm your host, Aaron Brooks. What does the energy grid and traditional it have in common? Well, both often struggle to keep up with the pace of change and the demand being put on them. However, for it, we have seen the move to cloud change the way that technology can be used in a more efficient and agile way, allowing for innovation to occur, to give birth to new ways to use applications and data to change our world.

[00:00:45] Aaron: Well, have you ever stopped to consider the state of our energy grid? If you lived in California, I'm willing to bet you have.

News clip: California has issued a state three energy alert, one step shy of ordering rolling blackouts for its 40 million residents. The state also reached a new record for energy usage. The heat add stress accrue battling more than a dozen large wildfires.

[00:01:05] Aaron: Every summer, millions of Californians experience rolling blackouts that disrupt the economy and poses serious threat to people's lives. It is a stark reminder that our energy infrastructure is not only fragile, but in many ways it's failing us.

[00:01:19] Aaron: The need for energy, agility, and efficiency is critical just like it is for it, not only for the sake of modernization, but in this case also for the sake of our planet. Enter Recurve a company that is pushing the boundaries of what's possible in the world of virtual power plants by gathering TROs of smart meter data and leveraging an open source platform called Flex Market.

[00:01:42] Aaron: Recurve is empowering utilities and businesses to go to market with green energy solutions that are not only sustainable, but they're profitable. Today we're joined by Matt Golden, the CEO of recurve, to explore the challenges and innovations behind his company's ground-breaking work. We'll delve into the importance of demand side flexibility.

[00:02:02] Aaron: And the complexities of modernizing a grid. We'll also look at how Google Cloud is enabling Recurves mission to accelerate the transition to a carbon-free economy. Matt Golden, welcome to the show. Thank you. This is a very interesting conversation. I actually spent a lot of time growing up, and as I mentioned in a, in a previous, uh, podcast, my father's a negotiator and he does it for oil and gas, oddly enough, and so my whole life has been spent.

[00:02:26] Aaron: Traveling North America to different hotspots where there was power plants or coal production. And we've always talked about supply and demand and energy grids. In my house, just as a kid, probably not the most exciting topic, but I was surrounded by it and it's amazing to me how much this topic has changed over the years.

[00:02:43] Aaron: I remember being a kid and it was all about turn the lights off when you leave the room or you know, shut the door because you're heating the neighbourhood. And then we've evolved to when I moved out, really understanding peak usage times and making sure I was doing my laundry at night to save money. And now we got smart thermometers and smart thermostats and home automation.

[00:03:03] Matt: Uh, and believe it or not, there are times in the day when leaving your lights on saves us all money.

[00:03:07] Aaron: That I need to dig into. Well, welcome to, welcome to the New World. It's a world that not a lot of people understand. So, very excited to have you here. I'd love to hear a little bit about your passion for this space.

[00:03:17] Aaron: Because you've been in it for a long time. You're clearly an expert in energy. Uh, what's got you so passionate about this? And tell us a little bit about Recurve.

[00:03:26] Matt: Yeah well, I'm a technology person first, you know, so I did a tour of duty during the.com, worked on a whole bunch of different just straight tech platforms, but really just found that I wasn't personally getting much out of it, you know, so I, but I, uh, I happen to believe climate change is real and it's coming faster than we all expect, and I have children and want to be part of a solution to that.

[00:03:45] Matt: And so that's really what underpins the company, uh, that desire to have a large scale impact to. Help decarbonize the grid and really got into this space initially through the solar industry, but version 1.0 of Recurve, which was also called Recurve, but a totally different company was actually a construction company.

[00:04:03] Matt: It was both construction and software going out and retrofitting buildings. And really, from that, learned how difficult that business actually is. But that's the last mile, right? And enabling, you know, the ground game, the, you know, primarily like Main Street businesses to scale. To go address the 130 million plus buildings in the US that consume energy.

[00:04:26] Matt: And you know, you're talking about the energy industry broadly. You know, it's the most complicated machine it often, which it is. It's huge. And, and, but it's also like, It's, it's intensely complicated and also extremely simple, right? We are trying to match the demand for energy with the supply in each hour.

[00:04:45] Matt: Actually each minute, each second, and each part of the grid. We have to keep the lights on. That's what the entire grid's about, and. Historically, it's been rather hub and spoke and you could put it up on a CRT in a control centre and turn things on and off, and it was still very complicated, but vastly simpler than it is today where we have what are called.

[00:05:05] Matt: Often DURs, DERs, distributed energy resources. So we're replacing those like big coal and natural gas plants with solar and storage and wind turbines and how you charge your electric vehicle and all sorts of like little things. And a lot of those things are actually in buildings behind meters, right? So they're not, there isn't even telemetry.

[00:05:22] Matt: No one's got a meter on the side of those devices. And so Recurves goal is pretty simple. We think in order to enable the grid to decarbonize and do so in a way that is resilient, the lights stay on and also affordable. We need to both address the supply side and the demand side, and what that means is we need to enable the.

[00:05:42] Matt: Energy use to match supply, not just the other way around. And so that's what we're about is how do we enable, you know, things that we talk about as energy efficiency, you know, insulating addicts, smart thermostats, and also demand response, which is, you know, getting paid in emergencies to turn off those lights and, you know, avoid energy market.

[00:06:01] Matt: You know, we've, we've seen some of the stuff in Ercot down in Texas where it's $6,000 a megawatt hour, which is a lot.

[00:06:07] Aaron: So we, we've, we've evolved from my parents yelling at me to turn off the lights is what you're saying, but same concept, just same. It is the same concept. It’s just a bit more nuanced.

 [00:06:15] Matt: Um, and what's intro, and I mentioned there are times you can turn the lights on and save everyone. Money is, you know, in places like California, which is California's crazy, but we're actually. A good, a good way to look into the future. Not that many years throughout the whole country where we have more solar than we literally know what to do with.

[00:06:32] Matt: So in the middle of the day in California and getting worse every single year, we are overproducing solar to the degree that we are curtailing grid side solar, like turning it off, and we're actually paying other states to take electrons from us. Wow. So on, uh, a really hot day in the summer and increasingly periods throughout the year.

[00:06:53] Matt: It's actually cheaper from an everybody's standpoint, a grid standpoint to locally run your lights. Because the alternative is curtailing more grid scale solar and paying Arizona to take our solar from us. And that's the world we live in. That is so different from the, the rolling brown oats and the black oats.

[00:07:10] Aaron: We had, uh, in the, I remember air conditioning bands where we couldn't use energy when I was younger, and I'm sure it'll still happens today actually.

[00:07:17] Matt: It's highly connected, right? So on the same day that it's super hot and we're over jenning at two or 3:00 PM. The sun goes down last I checked pretty much every day, and unfortunately that coincide like the sun is going down right about at the same time, or right before we get a coincidental commercial and residential peak where we have people coming home and they're making dinner and we still have, we still have restaurants and stores that are open, and so what we're seeing is.

[00:07:44] Matt: It's, it's called the duck curve, which is not looking much like a duck anymore. It looks like more like the Grand Canyon or like some sort of dinosaur where the middle of the day demand is plummeting because of all that solar, right? And then it comes roaring back. And so the name of the game is use and store more energy when it is.

[00:08:04] Matt: Clean and cheap and sometimes free and use less of it when it's super expensive and we're burning in some states coal in California and natural gas peak plants to keep the lights on. So it's all about just flattening the shape. So Recurve, that's our objective, is to enable what we think of today as virtual power plants.

 [00:08:23] Matt: Which is both energy efficiency, which is like base load. It's like, you know, you can't, you're not turning it on and off, but it's like hbe. You can't do some of that with hydro, but like, uh, nuclear wind, which varies, but you know, you can't dispatch it. You can't just turn it on when you need it. And also demand response event driven, which are like peak power plants and all of those roll up to a lot of activities and technologies in your house.

[00:08:47] Matt: You know, do you, if you insulate an attic, you're reducing your air conditioner load, which happens to be coincidental peak. If you put a smart thermostat on your wall, you can also shift and pre-cool when it's cheaper and better for the grid and coast through some of those expensive periods. You can charge your car in the middle of the day. Not during peak.

[00:09:05] Aaron: Right. And we just went through this in our, even in Ontario, we have grants, green grants that if you upgrade your window to a certain efficiency or you do change your insulation, it all ties into exactly what you're talking about, which is to help with that supply and demand and offset the carbon.

[00:09:19] Aaron: And I know you mentioned, um, the, your passion around carbon neutrality. So I just got to ask about how does this actually help reach decarbonization goals?

[00:09:28] Matt: So, I mean there's first of all, direct reductions by using less energy when it is high g h G. And so that's part of what we're doing and I'll, is measuring what's going on.

[00:09:37] Matt: Like you can't actually trade a product or value it if you can't quantify what just happened based on, I just put in a heat pump, water heater, or smart thermostat, or what was the effect of that. So one is just being able to measure direct GHG reductions, use less dirty energy, reduce the GHGs that you're, Emitting through your consumption.

[00:09:57] Matt: The other part is less direct, but is through this load balancing, which is, you know, supply and demand works in both ways. Like being, having too much, too many electrons is bad. Having too few is actually probably worse because blackouts are no good politically. People get, you know, it's bad for everybody on a whole bunch of dimensions.

[00:10:13] Matt: Um, so it's also just enabling through things like storage when you charge your EVs, but again, insulating your attic is also a blow reduction resource. And so it's about flattening that load chip and what you're really doing is enabling more renewables. Right? So, got you. You know, we need lots and lots of solar, but in California right now, as an example for the country and the, the globe and, but also in places like Germany, we, we have too much of it.

[00:10:37] Matt: So if we want, we need more of those clean electrons, but we have to be able to put them to work.

[00:10:41] Aaron: Yeah, I got you. So, so it's smart use of our energy and ensuring that both supply and demand are equaling the types of energy we use at different times. And that's going to help reduce the carbon emissions because we're using it more intelligently.

[00:10:55] Matt: I, and it’s not about saving energy, it's often about saving energy. But for example, electrification of buildings, heating, cooling. You know, water heating, cooking, and also vehicles is critical. Cause of course, you can't decarbonize fossil fuels right at the end of the day. So in that paradigm, we're actually increasing electric load.

[00:11:15] Matt: And if you do that carefully, you can actually increase electric load and decrease GHGs at the same time. You know? So again, it's, it's really about when and how you're using it, not the volume anymore. I got you. And there are actually good reasons to use more. It spreads the cost over a bigger number of electrons and it improves equity.

[00:11:32] Aaron: Actually using more is okay in a lot of cases. Got you. So I'm going t leave my lights on. That's what I've heard. Okay. Got you. 

[00:11:37] Matt: You got leave your lights on crank air. Um, I'm not really advocating that, but at the same time sort of like, honestly like, yeah, yeah. Charge your battery. Charge your car when we have more electrons than we know what to do with. And that is a good thing.

[00:11:47] Aaron: Right. Gotcha, gotcha. Part of that's got to be education, right? Like people don't know, they're not educated on when to do certain things and the impact of it. And I know that a lot of governments around the world have a lot of net zero deadlines, and I know it's just a few years away.

[00:12:00] Aaron: Uh, he is going to ask you a personal question. You, you know, it can do it off the record. Although this is a podcast, so it's not really makes any sense. You're going to be on the record. Are you an optimist or a pessimist that we're going to actually get that done?

 [00:12:12] Matt: I'm, I'm an entrepreneur. I'm a huge optimist. That doesn't mean I think we're going to hit these goals because the math doesn't quite agree.

[00:12:19] Matt: But I do think there's a tremendous amount that we can do. And frankly, I think we have the tools to do things a lot faster and better if we can get out of our own way. I mean, one of the biggest challenges we have in our very regulated energy industry is these rules have changed and they've changed fast, and we have new data in the form of interval meters and smart meters, and a lot of our biggest blockers are self-inflicted.

[00:12:40] Matt: Right. This is like we're kind of in purgatory between the old system, which frankly was a lot like, we'll give you a rebate for a water heater and whatever, and a new world where. And this is really where Recurves go comes in, which is how you utilize energy and like again, what we think of as a virtual power plant.

[00:12:56] Matt: And actually those three lines, and we're not, you didn't see my logo, but we have three lines in our logo. We're not just copying Tesla. That's, that means identical too. Right? Right. And the value of flexibility behind the meter is identical to a power plant. When you can measure and quantify its impact, but it actually is better because one is, it's hugely plentiful.

[00:13:14] Matt: You don't have to go into an interconnection queue or build new transmission to tap into it. And it also has this benefit of helping customers mitigate the cost of energy and climate change in their own buildings and make them better and come more comfortable. And all sorts of tertiary and ancillary benefits that come with it also.

[00:13:30] Matt: And so that's really Recurves role in our goal is to make virtual power plants identical. To those power plants on the other side of the equation, the physical ones

[00:13:42] Aaron: makes a lot of sense. It actually, uh, almost sounds like what we deal with today in  it around on premise data centres and then the cloud.

[00:13:49] Aaron: And when we talk to customers about cloud and how they leverage it, it is very utilitarian. Where it's a utility model of consumption and you have to make sure that the applications that you have and the data you store, as you would know because you're a technologist, um, you have to rely on those resources.

 [00:14:06] Aaron: And if you over resource, you're spending too much money and it's inefficient. And so there's this elasticity in the cloud that seems very similar to a plants that you're speaking of and managing that most effectively is where you're going to see the best benefit from. Performance perspective, a cost perspective, and an environmental perspective.

[00:14:22] Aaron: Now, I know you've done a lot of this in California, and I love these, the, uh, work that you've done with the largest utility there. Maybe you could share a little bit about the work that you've done in California, uh, in terms of how you're helping them and, and the impact Recurve has had in the Californian market.

[00:14:37] Matt: Yeah, absolutely. And first of all, thank you for that new analogy. I'm going to steal it liberally. Uh, it does, it is, it is very similar. Frankly, when we first started, we're doing this type of work and we work like we're exclusively in the cloud, which is necessary because you know, sometimes we have to throw 30,000 servers at a problem.

[00:14:53] Matt: But when we first started, utilities looked at us sideways when we said this is software as a service and we're down to one utility I'm aware of who won't allow cloud data infrastructure. So, and that's okay. We're seeing that change. So the work we're doing in California and really across the country.

[00:15:09] Matt: You know, as a smart up, you got to go where the pain is. And like, the pain is the most acute here, though interestingly, miso, which is like squarely in the middle of the country, has the most imbalanced grid, not California or New York. So these are not just California problems, we're just the craziest. So ours are the biggest, you know, we do things first and fast, and then we sometimes learn painful lessons.

[00:15:30] Matt: Sometimes we don't. Again, from our background and as a company, it's really our, our business is three, three simple chunks really. The first is, Before you can value something, you got to actually be able to measure it. So that's really where we started was how do we use interval data to measure. And, and the problem is, measures even the wrong word.

[00:15:49] Matt: When you change a building, and this is the, an interesting one that is pretty obvious once you think about it. But if I wanted to measure what a smart thermostat did to your energy use, when you install that smart thermostat, there's no building left that doesn't have one to measure against. So, so we can't measure, we calculate a counterfactual.

 [00:16:06] Matt: What would your building have done into those same conditions if you didn't have a smart thermostat? Interesting. That's the interesting conundrum. And it turns out, here's the problem we ran into is if you ask 10 engineers what the savings are on the same thousand buildings, you'll get somewhere between 14 and 16 answers all correct.

[00:16:23] Matt: But they might vary by 40% because there's hundreds of little engineering decisions that are being made that we haven't agreed upon. So step one in our business was, this is not art. This is math and science. So we're going to run an open source process to develop a standard set of measurement methods and code and make that open source for everybody.

[00:16:43] Matt: Nobody can own a kilowatt hour, a pound, or a bushel of corn and then trade. And that's really one of the more disruptive things we've done. Um, and so what we've ended up with is, is a bunch of tools. You know, the important ones are they'll all eventually be in Linux Foundation, but they're really critical ones are, you know, we don't own, they're in the Linux Foundation because again, we don't think you can build a market on something that's a black box, right?

[00:17:04] Matt: So first, the ability to forecast and measure the impact from that smart thermostat, that HVAC system, that heat pump, whatever, it's, and know how good it is. We call these revenue grade calculations because one, we have uncertainty and confidence intervals that we can place around these things. It's not about being perfect; it's about being able to discount properly.

[00:17:22] Matt: And the other component is being able to put it in a contract and have, again, get rid of that issue. You know, anybody that runs the code the same way will get the same answer. Doesn't matter who you are. And if you want to have a better way to pick weather stations, we have a process, actually an ongoing process.

[00:17:37] Matt: We have working groups, one that met this week to improve those methods over time, but, you got to make an argument and prove it's better and then then sure. So that's the first part of our business to be able to do that. And we are able to take that type of analysis and because of the cloud, rather than just have an engineer do m and v, you know, an Excel spreadsheet or locally on that, on your building, we're able to do it on every single meter, on the grid in your real time, you know, thanks to.

[00:18:02] Matt: 30,000 servers, we can throw out a problem, you know? And even then it might take three days to run all those meters. So that's step one is commoditizing the ability to measure and making it revenue grade and consistent. And having a standard definition for it. Exactly. So everyone's on the same language.

[00:18:17] Matt: Exactly. Now there's a lot of things you can do with that. Like we definitely know who are the people with bad air conditioners and. Who would actually save money with the heat pump. And it's remarkable how you can pinpoint this information right at the grid level. So there's a lot of use cases around planning and forecasting and targeting and the right customers that'll have good outcomes.

[00:18:35] Matt: But the reason I care about it is that you can't trade unless you have a product. And that product is a calculated value. So the, the next part of our business, and this is really what we're doing. Ain't rocket science. Instead of picking a winning technology, we let companies who have solutions for customers sign a flexibility purchase agreement and agree to get paid based on what actually happens.

[00:18:57] Matt: We can value what is the benefit to capacity, long-term load balancing. What is the reduction during events if we've dispatched. What is the transmission distribution and GHG value something in California's called Total systems benefit. And we pay those aggregators for the value of their portfolio, their virtual power plant, based on the total systems benefit they deliver, which is kind of just the way everything works at the end of the day.

[00:19:20] Matt: And that's different than your rebate program, which is we're going to design a rebate, pay you in advance, and hope it actually works, which really misaligns all the incentives. The incentive here is these aggregators have a strong incentive to. Deliver solutions that actually work or they won't get paid.

[00:19:33] Aaron: Yeah, so it's the, it's the measurement of that. It's, again, I'm going to go back to the cloud analogy, but you, you're, you're paying for what you use. It's the same thing with subscriptions. Even today, when you think about SaaS outside of what you're, what you're doing today, if I'm not using a certain feature in.

[00:19:49] Aaron: Uh, it doesn't matter what I'm using, what the tool is, I don't have to pay for it. I can turn it on and I can turn it off. And so there's a real sense of control there. But in order to do that, in effectively, bill and see benefits on both sides, you do have to measure effectively and you have to be transparent with that measurement so that both parties can make sure that they're, you know, not getting kind.

 [00:20:09] Aaron: Hosed if you will not, you know, they got to believe it and it's got to be truthful, right. So they got to see it. 

[00:20:14] Matt: And it’s also important that these customers, right, you talked about education well to a degree, but like the volatility in the grid, you know, your, your end customer in a house or, or even like, even your tip, your Class A buildings, very few of them even really have energy managers, so they, they're not in a good position and don't have the power of large numbers to really be exposed to energy market price volatility.

[00:20:32] Matt: And so the other is like, you know, people are buying cold, cold beer and warm showers, not t and d for the utilities. So we also also working through a layer of aggregation. And this is not dissimilar to us working with, with your company and actually to help us manage ours. You know, Google costs and other things, which is we send a really sharp price signal to the aggregator who has, you know, thousands of projects and customers, and some of 'em are great, some of them don't, and they have the power, large numbers to manage that.

[00:20:59] Matt: And then they will translate that into something really simple for the customer that they can actually grow.

[00:21:05] Aaron: Well, I got to tell you, I love hearing how Recurve is shaping the future of energy and how they're using technology to transform their industry. And that is what we see time and time again from our customers across all industries.

[00:21:17] Aaron: I wanted to know how Recurve was using AI to tackle data collection challenges, but before we hear from Matt, let’s have a quick word from our sponsor. Today's episode is brought to you by Google Cloud. Are you ready to make smarter decisions with your data? Well, Google Cloud provides a unified data and AI solution that's really unmatched in speed, scale, and security.

[00:21:38] Aaron: All with AI built in. Companies like Recurve are leading the way in using and scaling their data capabilities on the Google Cloud platform. But even if your data is messier than your kids' bedroom, Google Cloud and Softchoice can help reach out to Softchoice and inquire about joining one of our data analytics accelerator workshops.

[00:21:57] Aaron: We'll help you break down silos across your organization's operations, analytics, and data stores to drive new insights and innovation with Google Cloud. Don't wait. Take your data strategy to the next level with Google Cloud and Softchoice. So it's, its data collection turned into some form of reporting or insight that then gives the ability to manage this effectively by both parties with being very transparent with common language.

[00:22:25] Aaron: I'm curious, as you mentioned cloud a few times in terms of how you're building this out so you can throw compute power at it. Have you explored or are you doing already anything around? AI or machine learning as it relates to the correlation of this data where it's actually getting proactive and seeing ahead of time versus being reactionary?

[00:22:44] Aaron: Like how much of your business is that?

[00:22:46] Matt: Well, that's actually, that is a, a major area of our business that we're focused on now. So the core algorithms are not a good match for machine learning or AI from the standpoint that. We need to be able to get the same answer over and over again. And frankly, it's important that I could explain in an intuitive way what is really going on, which of course is a challenge with AI.

[00:23:05] Matt: Um, but also you can't train. And so the other thing, by the way, is, is often a thousand times more computationally intensive, um, right at the same time. And some of the stuff we've been looking at doesn't necessarily produce a better result. So on the measurement, we're really like, you could, frankly, we're doing.

[00:23:18] Matt: Absolute tons of regression models, like I think we would actually might fit into machine learning out, but it's, I wouldn't consider machine, but it ain't that different. Cause what we're really doing on those tens of thousands of servers is just insane amounts of regression modeling. So we, we might run on the same servers even, but it's not really AI where we can also really do a lot with machine learning is, our first part of our business is agreeing on standardizing and commoditizing measurement.

[00:23:41] Matt: The second part is enabling the market to get paid for their virtual power plants. Utilizing that standard measurement. The part that we're a lot focused on today is how to do two things, really. Forecast grid impacts so that they can be integrated into the long-term forecast and planning on the grid, or you're just kind of yelling into a black hole.

[00:23:59] Matt: Interesting. And also what we're doing for the customer is really cleaving off the benefits. If I buy a heat pump for my house, like I'm getting an air conditioner, I might get lower bills. I got to admit, I bought heat  pump, large part because it gets hot and Smokey in Mill Valley where I live. And you need these days when it's hot and smokey and I can't open a window, it can get pretty miserable.

[00:24:20] Matt: That was what might be the most important outcome of a heat pump to my family. That's what I bought. Right. But it can have impacts to the grid. So we're cleaning off that the virtual power plant, and we're saying we're going to pay for that over time. Right, as of just like a power plant, right? Which means instead of the customer paying for it with credit or asset value or cash, you know, we can finance that like an infrastructure investment.

[00:24:40] Matt: So, if a bunch of companies are going to go out and retrofit thousands of buildings and get paid over time, being able to forecast and underwrite that cash flow. So it can be financed in the same way as your power plant that your dad used to build, huh? Right. Huh. Which is based on the long- term cash flow in project finance or project finance and infrastructure investing.

[00:24:58] Matt: We'll say a customer is doing project finance, which is really not true that you're doing consumer finance, because you got to, it's not what you finance, it's how you financed it. So financing cash flows from virtual power plants is an infrastructure investment. And that's project finance.

[00:25:11] Aaron: Sounds like there's a lot going on in your space, especially on the technology side.

[00:25:15] Aaron: What's the biggest hurdle you've had on the technology side? Like when you think about your move to cloud and the the need to scale up and scale down in order to support the work that you're doing, uh, what is your biggest challenge, Ben? Bad data. Bad data for that. Pretty much from every customer I've spoken to, you know, as a ETL is 80 per cent of data science, right?

[00:25:35] Matt: So yes, absolutely. One is a lot of the data we're dealing with is. And we believe this to be important is PII and there is customer information in there. So really being highly secure and also thinking very carefully about how we put the data to work that protects individual privacy and also from a grid standpoint and a regulatory talent, but you know, balances privacy with the imperative to decarbonize the grid and put this data to work because.

[00:26:01] Matt: If we can't put it to work at all, why did we just invest billions of dollars in smart meters? Right. Just the fire meter readers? No, we're supposed to do something with it and then we're constantly working on our computational efficiency and it seems like every month we set a new record on we call doing these baselines meters.

[00:26:17] Matt: So Yep. If I say like number of meters run, I actually, what I mean is like baseline calculations or. Forecastable, but you, you know, seems like we're constantly pushing, you know, pushing the edges on our capabilities to do this work and, you know, improving the efficiency. 

[00:26:31] Aaron: Speaking of that, like on the pushing the edge and, um, thinking about the future, like we've come a long way and I, we're starting to see innovation really get crazy with generative AI and other things that are just coming out to the woodworks.

[00:26:42] Aaron: I always joke like, I feel like I woke up one day and the entire world. In terms of AI has changed, and I don't know, did I? Was I asleep for a year? I have no idea how this just came across. I had the same feeling, so it's really weird. All of a sudden there's a thousand ways to do something completely different.

[00:26:57] Aaron: Now. When I think about your industry and the evolution that it's been on and your use of technology in this space, what does it look like in five years, six years, in terms of the impact that this type of focus on data around our energy grid and decarbonization will have on the planet and on people.

[00:27:15] Matt: In that last step of forecasting, that's where, you know, artificial intelligence and machine learning really potentially shines, which is, you know how to use this huge body of data to increasingly, accurately forecast loads. Using both meter data and also all sorts of app dependable data. You know, we know when the sun's out, we know much insult insulation you're getting, we have all sorts of potential to append all sorts of data about your building stock, et cetera.

[00:27:40] Matt: So it's in that forecasting component that I think that's where we're starting finally, to really put machine learning to work. You know, we have some interesting. ML models for detecting where smart vehicles are being charged on the grid, which 60 per cent of times really easy, but as you start to get to the edges, gets really hard actually.

[00:27:56] Matt: Right. So that, that's kind of where we're putting it to work. And I agree like ChatGPT has put everything on the accelerator, that's for sure. And yeah, it did feel to a degree it came out of the blue where holy crap, I had no idea that techno that we were at this place this bad.
[00:28:08] Aaron: Exactly. And I'm curious that like the impact on just society, like if I wake up five years from now and we've figured out some of. Our energy supply and demand concerns and we've decarbonized like, what, what does that look like? And how do you think that'll impact the, the everyday person that's just waking up in the morning, like, how does this look? Five, six years now?

[00:28:31] Matt: So that, that's also like why we're oriented the way that we are, which is like, we want this to be as minimally invasive to your average person who does not need to understand this complexity.

[00:28:41] Matt: Right. It is, you know, frankly, people are like, this is so complicated. I'm like, you use a credit card. Right. It's not a diff the back end of what we're talking about is like the same. Yeah. And in fact, that's our big innovation. I, I did a bunch of work when I sold my construction version of recur before I started this new version of it in like solar securitization.

[00:28:59] Matt: And frankly, as an energy person out, I, I'm often considered a finance person just because I know. You know, a little bit, right? You just have to know more than everybody else to be an expert. But I was involved in those securitizations, like on an asset management standpoint, this is where the money is interesting.

[00:29:13] Matt: So what we're really just trying to do is frankly, make flexibility look like a solar deal or a mortgage-backed security or a car loan. And the complexity of the backend around actuarial analysis and trenching of risk and everything is like, frankly, not a heck of a lot different than how your credit card works, but you know how to use your credit card, right?

[00:29:31] Matt: But, What's on the back end? You know how you go in and you buy a car and you get a car loan? Yeah, that was easy. What's on the back end? Same thing. Same ideas, right? Nothing. Nothing. All that different. So what I hope to see from a consumer standpoint, like first of all, like the thing we're trying to mitigate is.

[00:29:49] Matt: Anybody that tells you decarbonization is going to be cheap or easy is selling you something, it's not. And so at a macro level, we want to enable us to get to a clean grid while keeping it affordable and keeping the lights on. And then on a customer side of things, we want to see, you know, there's about 150 companies now in our flexibility Flex market provider network, we call them aggregators.

[00:30:11] Matt: Mm-hmm. They range from. You know, companies doing lighting in low stores and refrigeration and Albertsons to HVAC and low income customers homes, right? Like you name it. Well, what we'd like to see, and I'd like to see is all sorts of amazing, innovative business models coupled with technology that are providing solutions that make people's lives better.

[00:30:29] Matt: Frankly, it doesn't have to be passive, but has they have lots of options to help manage this risk and someone else to make it easy. You know, I often hold up, like there's a lot of talk of like real-time rates to customers and I'll hold up my phone and say, who has a cell phone in this room? All these advocates for real-time rates.

[00:30:44] Matt: Do any of you pay by the minute? Oh, none of you pay by the minute. Interesting, interesting.

[00:30:51] Matt: Yes. That's a great analogy actually. So like the idea is to make it as mini, give customers lots of choice and lots of options, and have companies competing to deliver awesome solutions to them.

[00:30:59] Aaron: It's very interesting and it's, um, we're seeing this quite often where you find it a good idea, somewhere in somebody innovate something in some industry, and it's like, how could I use that in other ways?

[00:31:09] Aaron: And so we're starting to see some of the. I'll call them older legacy industries that are really starting to modernize with technology. And this is a, a great example of, of just a, just another one in energy. L

[00:31:20] Matt: like there's all sorts of amazing clean tech out there, but there's also companies we work with, you know, managing refrigeration in beverage warehouses.

[00:31:28] Matt: That have, you know, stuff out of the nineties in net and control a tremendous amount of load and we just, you know, hey, did you know you could also be in the energy business and make money by creating? I didn't know that. Right. So it doesn't have to be, you know, energy is the biggest, right. 

[00:31:44] Aaron: It touches everybody for sure And so, and it, it's like, um, Google, when they talk about their technology, they actually don't talk about the speeds and feeds and the
 things and the stuff you can buy. They, they talk about and try to empower others like yourself to use their. Technology for an outcome. And so you're taking the same kind of approach.

[00:32:00] Aaron: It's like, yeah, you, my long-term goal here is how can I enable others? Watch out Google, we're coming. Yeah, that's right. Uh, I'll make sure we send this to them for sure. Um, so that was really great. I, I appreciate it. I know there's a lot of people that we have questions around. How do they as individuals that maybe want to build on this platform or take advantage of this movement that they can learn more?

[00:32:22] Aaron: Like how do they, where do you. Push them to, what do you encourage them to do in terms of reaching out to find out, um, more potential opportunities to work with you?

[00:32:30] Matt: Well, I mean, from a plug standpoint, recurve.com, which also links to a website called, which is demand flex market.com, which is actually, if you're, this isn't really for your end customer.

[00:32:41] Matt: Like we don't deal with end customers by intention. Like we're sure that would be a kind of a conflict of interest and like just not our position in the market. We wanna enable others. But if you, you wanna start a business. Helping people manage energy use and again, capture that benefit to the grid and not just, you know, we make it easy to do.

[00:32:57] Matt: Part of it is breaking down the barriers. You don't have to compete with billion dollar PE funded implementation companies for a contract, right. Where it's winner take all. And that's a big part of it, by the way, like giving people a ticket to ride. If you are, uh, a big nerd like we are, The open EE meter.

[00:33:13] Matt: If you type in open E meter, Linux Foundation, energy, l f E, you will find all of our code on GitHub. That's amazing. There's actually a, oh, there's a working group working on CalTRACK, uh, which is the methods, the open E meter 2.1. Which will roll right into 3.0 shortly, where we're making those methods better.

[00:33:31] Matt: But it's Python, it's open sourced, it's under an Apache two license. Have add it, please contribute back. Yeah, that's great. So that's great. You know, we want, we wanna see in that innovation happen, even if it means eating our own lunch to some degree.

 [00:33:45] Aaron: For sure. I love it. Competition never hurt anybody and crowdsourced the heck out of everything.

[00:33:49] Aaron: Uh, so, so listen, Matt, thank you very much and for anybody else that's interested, you can also Google Matt and the work he did with Ellen DeGeneres, I believe you, uh, went through her house and helped her there. So a little personal plug there. That was Reer 1.0, but Okay. It's still an interesting story that that was when we actually worked for a living in retrofitted buildings.

[00:34:06] Matt: That's great.

[00:34:06] Aaron: That's awesome. Well, thanks again, Matt. I appreciate it and uh, look forward to seeing where Recurve goes in the future. Appreciate it. Thank you. Well, we're at the end of another episode. A big thank you to Mac Olden for joining us, sharing his insights on the importance of demand side flexibility.

[00:34:22] Aaron: And modernizing our energy grid from power plants to open source platforms. It's fascinating to see the impact technology is having on the future of this industry and to see what companies like Recurve are doing to lead the change impacting people business and our environment. I'm excited to see what the world looks like in a few years.

[00:34:43] Aaron: Thanks for listening to The Catalyst, a Softhoice podcast about the amazing things that happen at the intersection of humans and technology. I'm your host, Aaron Brooks. This episode was brought to you by Google Cloud. See how Softchoice can help you take full advantage of the cloud with our data analytics accelerator workshops today.