Energy Tech Startups

0:00 Transform your startup journey with the energy tech nexus. Connect with fellow founders, access critical resources, and be part of a community shaping the future of energy and carbon tech. Your

0:10 path to building a Thunderlizard starts here. Learn more at energytechnexascom. Welcome to

0:16 another episode of energy tech startups. Today, we're doing a little bit of a different format. I am here in the studio with Jason, who's usually with me, but I'm also bring We're bringing back a

0:27 guest, a Sujata Kumar, who is the CEO and founder of Decider. And since we're bringing her back on the show, instead of just going over the same questions that we went over last time, we thought

0:42 we would do a little bit of a roleplay. This time around,

0:46 a Sujata is bringing Decider, a building Decider

0:52 She really wants to now see how this tool could help.

0:57 founders that are working in climate tech and heart tech and how the solution that you're building with Decider can benefit a founder. And so Jason here is going to play the role of a founder. Since

1:09 he has been a founder, it's still an entrepreneur today and really has that experience of having worked with many founders within a clean energy, climate tech, heart tech, whatever you want to

1:19 call it. So to start off with, I'll ask you a question Sujata, to begin with, explain to us a little bit

1:32 what you're building and why is this solution - why do you think the solution is going to be really important for a founder within climate tech, energy tech? Absolutely. So Decider's key remit is

1:46 to deliver two things. One is techno-economic analysis, or TEA, and integrated planning. And why is it important? It's really generally important. whether you're doing clean tech or not, but

2:00 clean tech, it's more important because the complexity is high and we don't have prior data to actually, you know, kind of go out of like, you know, if you've done. If you're doing oil and gas,

2:15 you have multiple years of data to go by and say, I've done this before, this is how it's going to happen In clean tech, this is all new. Everything is, you know, completely. It's, you know,

2:27 first of a kind in almost every situation. And so how do you actually gain a full understanding of what you're building, both from the efficiency capacity perspective? And how does that relate to

2:41 economics? If you don't bring that together and kind of think about it in its entirety, we call that system model If you don't, then what happens is, you know, the sustainability or the adoption

2:57 of it. will not happen. So that's why it's important to actually bring these two together, both the efficacy of the technology and how much emissions is it generating or rebating and bringing it

3:12 with

3:14 economics.

3:16 Can you talk to us about what kind of an energy tech company would this be relevant for? Yeah, since we're doing hard tech, I mean, since we are doing techno economic analysis,

3:29 it will be primarily hard techs. And so the way we build a cider is it's really connecting the ecosystem. On one side it is all the hard tech, you know, think about companies doing point source

3:45 capture, direct air, electrolyzers,

3:49 turbines, solar, micro grids, all of these companies are trying to innovate and bring together new technologies. And they have to work with investors, and they have to work with customers. So

4:05 think about how all of these companies, the customers and the investors, whether they be banks or venture capital, they all have to kind of think about, okay, we are investing in these

4:17 technologies, or we need to adopt these technologies into our environment. How is that going to work? What are the, how am I going to reduce emissions? How am I going to, how is the economics

4:29 going to play out in the next five years, 10 years, 20 years? So that I can continue to reduce carbon and create those low carbon products. And these, you know, the venture capital has to look

4:43 at it and say, can I, I mean, is this true? Like, is this technology really going to do what it's saying it's going to do? Is the techno economics actually real? So it's, Decider actually has

4:55 a fit for purpose for all these. three groups in the ecosystem connecting them all together and you can start as a starter. Just say, hey, I need to know my own techno economics, but I also need

5:10 to share it with my VCs and I also need to share it for pilots or, you know, just adoption in the customer and the industrial companies, be it oil and gas, utilities, you know,

5:27 chemical companies, whoever, all these companies have to, you know, look at. So for a startup that's building a first-of-a-kind industrial project. Yes. And that's you, Jason. And if I can

5:37 kind of jump in. So like, you know, the first time I was building a company, I didn't know anything because I came out of college. I don't think I took an economic, well, I took an economics

5:46 class. I didn't do a financial modeling class And so we get thrown into this world where, you know, the questions come up. And I think every entrepreneur kind of goes through this journey where

5:55 they realize they have to get more sophisticated because. your first answer is never good enough for customers or for investors. And, you know, I think in a classic journey, you come out and say,

6:05 oh, I got this brand new widget. It's gonna save everyone money. I should just put a 30 margin on it as an example. And maybe that's actually not a good margin, but you know, that's probably

6:15 your first thought as an entrepreneur is like, I'm gonna do cost-based pricing, right? 'Cause I don't really know what it's worth.

6:22 And someone just told me to do that once. Like that was, that's like the first thought process like how I priced my product the first time. Terrible idea for a number of reasons. One, I didn't

6:31 really understand what was valuable to the customer. And so being forced, and it made it hard to kind of communicate to a customer because when you go to them and you just tell them, oh, it's

6:41 cheaper, well, it isn't really valuable. It's just, you're cheaper, right? What you really wanna do with a techno-economic analysis is to jump in and say, well, I really understand your

6:51 problem as a customer. Like this is really your pain point and what your pain point costs you.

6:57 So that's either time or it's labor or it's materials or sometimes it's more sophisticated. It's downtime or it's the loss of something and you can kind of price your product based on the value

7:10 you're creating for someone and in many ways you want to share value with your customer, right? So it's not that you're reducing their initial purchase price by 90, it's, you know, we're giving

7:21 you something that you know is valuable because it's based on your own analysis It's based on how you spend money today or how you view your internal economics. And when you can kind of align your

7:33 pricing or your pricing model with the same language they speak, that's when you have really good kind of product market fit and really good alignment. And the techno-economic analysis is kind of

7:43 the storytelling. And I think for a lot of founders, the first time you put together a financial model, you don't appreciate that you're really telling to tell a story of the future and put it

7:54 numbers so people can see. Okay. There's. There's value being created, there's value being extracted, there's savings on all sides, you know, and in a TEA, you gotta put in OpEx. Well,

8:05 what's the story of telling? It's telling a story about maintenance and consumables and all these little things that you have to think about that provide confidence that the value is truly there.

8:17 And it makes, it turns the

8:19 story from a fairy tale into something more concrete. There's nothing more concrete than actually doing it, but this is the next best thing as we can kind of create a simulation of what life is

8:29 gonna be like, and that all gets kind of wrapped up in the TDA. Yeah. I mean, Jason, you pulled the words out of my mouth. It's really storytelling. It is a roadmap to their success. And you

8:44 cannot,

8:47 by simulating it, you're actually creating a version, virtual version of what it's going to look like And that is going to. iterate over time. So how do you build something that you can really

9:02 just think about how is it actually going to work when it's executed? And try all these different scenarios, something you can do in digital technologies that you can't do in real world. It gives

9:15 you a blueprint of what a real life plant is going to look like, or when your technology goes into a particular environment. What is it going to look like? And what it's you can do is fundamentally

9:30 important for startups. You brought up something else that's important, which is most of these -

9:38 the founders that are coming out, they are scientists. They are brilliant people. I mean, this is not easy to innovate. All these technologies that have been brought out. So now you are asking

9:49 them to do something which very, you know, complex, which is, oh, go into the, the, the granularity of how these metrics are going to be. They know how this technology is going to build. They

10:02 are innovating on the technology, but then there is something, a wrapper they need to put on it, which is all around, hey, if this happens here is an yield here, if this happens here is how the

10:13 economics are going to work. Oh, by the way, the IRA is putting all these different rules in. How are you going to bring that into the equation? How are tax credits going to happen? What are

10:25 subsidies going to happen? If you go into Europe, how is that going to work? If you're in the US, how is that going to work? How do you plan for energy? What is the price of grid? Just in a few

10:36 minutes, I have just articulated the number of parameters and data points they

10:44 need to consume in order to do this techno-economic analysis. One of the things we struggled with - when we were building our gas turbine technology, you're trying to service the shell market. And

10:57 we were converting Flaircast to power. And this weird thing happened in 2014, I think it was 2014, when they started to go from single pad drilling to multi pad drilling, and then the wells just

11:09 got big. And so there's just the power demand shifted significantly. And so we had gone to investors, we got into customers and say we're gonna build a widget that's size one. And then almost over

11:20 18 months, the size requirements went up like six times And we had to rebuild the models. And it was almost from scratch. And it was such a big question of like, what's the right model size to

11:32 build? And you almost need to do a trade study of how did the economic shift? Because even though we get bigger and we wanna go off and build a bigger system, the LCOE for our competing, the

11:46 competitors which were regular gen sets got better. And so we were kind of in this environment where we needed to build a bigger product, It's going to be more RD spend. But it was unclear if we

11:59 would have as much of a competitive advantage. And so we had to almost like rebuild the TEA again, 'cause we had to really understand what the competitive landscape was, 'cause that influenced the

12:11 value we were creating for our customer. And I think just the increase in size decreased our value by like 50 in terms of the LCOE of our competitors Luckily, there was some scaling rules, so us

12:26 getting bigger actually had positive impacts on our own economics. But it was an important question when we're trying to decide, are we gonna put RD dollars into building something new and bigger or

12:36 not? And how do we maintain that kind of flexibility? And you almost wanna

12:43 pick a lot of scenarios. It's not like, is it better to be five times bigger, six times bigger, seven times bigger? We needed to kind of understand like how those levers shifted, I didn't know

12:53 what was going on in this jail industry. I don't know if a lot of other people really had that foresight or maybe the guys who had foresight are a lot wealthier than I am, but I didn't see a lot of

13:02 those changes coming. Um, but having that like knowledge about how my strategy to change really required a model that, um, you know, I had to rebuild an Excel and it was, it was not that it was

13:15 painful. It's just, it was hard to articulate what was driving pieces of it. Um, cause a lot of it ended up living in my head Um, and not necessarily didn't necessarily make it into the story the

13:26 right way. And, you know, think of it different ways, right? So TEA, yes, you know, we talked about it. It's, it's, you know, if you do it correctly, you need to be able to scale it. But

13:37 also you need to be able to optimize it. So I want to get an LCOE of X. So what should my price be? What should my capacities be so that, you know, how much power consumption them having so that.

13:53 And now you can just look at it and say, okay, if I want my LCOE to go down, what all should happen? So you can get a recommendation if you do it right, and you, you know, you can look at it

14:05 and say, oh, I, this needs to go down, this needs to go down, this needs to go down. Can I do it? Where is my RD going to be? That's the development plan. Exactly. Yeah. And, and so that

14:17 is the benefit of it. And it allows you to scale, you know, you can start with small and say that you're the first scenario. That's my baseline. Now I need to evolve. And now I'm going to just

14:29 scale it. What's a comparison? What have I changed? How is the metrics changed? Do scenario analysis and just do a compare scenarios and say, oh, this is how I've changed it. Yeah. And, and

14:42 those are the benefits of doing that, you know, in a very structured way And,

14:49 you know, going back to, you know, Adi. Innovators going to spend their time building out their technology in a robust way While you need to have you know Support in trying to do this research and

15:05 bring this analysis to you so they can just go in they know what they know How can they enter certain parameters and say I know

15:23 this is how I can make my technology more robust and then plug it in and they see the outcome It's really that's where that shift has to happen into a TEA making it easier and easier for founders to

15:29 do what they are doing very well and Make the other aspects of it simpler There was a founder who actually told me this many founders actually say this to me They said the biggest thing that they find

15:43 is they have to share this analysis with They're investors. So go to a VC and And.

15:51 they'll share their Excel spreadsheet. And then the Excel spreadsheet is an Excel spreadsheet, right? So they will, you know, that's baked on real assumptions. As scientists, they have put the

16:02 real assumptions which comes from their experience, scientific documents, et cetera. The VCs will go in and change a certain parameter and come out with a result and say, Hey, your answer is

16:14 wrong.

16:16 And so these guys have to go back and say, what did you change? Like, how could it be? So they come to me and say, if they had a software which they can just say, here are the parameters you get

16:28 to change. So this was a classic thing. It was probably my fourth financial model where I realized I had to create a parameter sheet. Yes. And I had to lock everything else. Yes. And it's so

16:40 mind numbing. Meaning like I didn't like having to like, every time you'd make a parameterization, then you'd have to link it from the sheet somewhere over here and sell AB to, you know, the,

16:51 the, the, the sheet on the other side, and you really had to think through even what the ranges would be. And there's no validation, I think, on Excel that can control that. But no, that was

17:03 like, you had to really go through each line and think, OK, what is something that someone would want to vary? It was

17:12 like an argument we would have, do we want to vary interest rates? Which can affect the TEA a lot. A lot, yes. Your net present value And it's like, what are we going to estimate the risk-free

17:23 interest rate? I don't know. But it's one of these things that can really drive your TEA. And it's got this funky equation you have to implement everywhere. So those were little things that you

17:35 would have to think through. I think the other thing that I was never able to do very well was to do a tornado analysis where you have to go in and see essentially where I'll break down if like the

17:50 parameters change. You probably know what a tornado is better than this. So explain what a tornado is. Yeah.

17:56 Maybe you don't. No, it's like a sensitivity plot. Yeah, so the sensitivity plot, you can do it in, you know, a tornado chart or, you know, we do it like we can do a Monte Carlo simulation.

18:08 Ooh, fancy. That's even more fancy. Yeah, so we can do a Monte Carlo. Now that's, you know, pretty difficult because, you know, you can put an optimizer on and you can select which parameters

18:19 you need to actually vary to actually get a result that is more optimized. You can optimize on an NPV, you can optimize on emissions. But then doing a Monte Carlo will sample different data points

18:35 and it'll give you a chart which says, okay, you're at the basic sampling and you can just go in and see where the risks are. And that will give you kind of a, you know, a very good, good robust

18:50 analysis that most VCs would understand. I mean, think about it, your mutual funds run on Monte Carlo, right? So, you're looking at your future, looking at, okay, if the stock market

19:02 performed in these different random points, what is your retirement account going to look like? And we are doing the same thing for

19:13 technologies. So companies, customers or VCs can understand, what is the baked-in risk that we have to assume in this? And

19:23 what needs to change to alleviate those risks? Yeah, I think

19:29 with our company, like one of the things we failed with, to be honest, was we never had a good way to explain

19:38 how much reliability mattered. And from a customer perspective, downtime is very expensive and

19:47 You know, we did sort of a sensitivity analysis on the value to the customer, but Amani Carlo would show you just, you know, an extra, I think four hours of downtime for our customer a month,

19:58 would translate to like a2, 000 loss, which was the entire, like, like to pre, it was, that was more expensive than the OpEx and the depreciation on our hardware. Absolutely, absolutely. And

20:10 it's like, well, how do you mediate that? Well, you need to provide an MSA, you got to provide service, and as long as your service is cheaper than their cost, it would be super valuable, but

20:18 I can never kind of get the math to show up where I could

20:25 convince whoever I need to convince investors that we need to be capitalized to do that properly. And because we, 'cause I wasn't able to raise the money, we actually bombed a pilot because we

20:36 weren't able to provide the kind of customer support we needed. The customer never came back. He was just too expensive for that. You get one chance. We got that one chance And we just - just

20:46 didn't allocate kind of the resourcing that we should have. And I think that's a, you know, wiser now. So I wouldn't make that mistake again. But it's just one of those things where the customers,

20:57 and everyone around the customer would say, you need to be there, you need to have our backs, you gotta provide that kind of extra level of service. And we just didn't appreciate how expensive it

21:06 was. So if you had a techno, if you had been a techno-economic analysis where you could have said, okay, I'm gonna do it's you so much downtime for your, all right? That's going to give you a

21:19 better - Let's just on our pilot. Like this is the pilot unit of, we didn't put into the budget, enough aftermarket support. And so we blew through the aftermarket support budget, like within a

21:29 week, and I was out of money. So I couldn't actually provide more, but that's what made the first of a kind of pilot fail. Was not like, the system ultimately got running and we got it working,

21:40 but it had already kind of poisoned the mind of the champion.

21:46 After we had been running for like six months, they said, well, we're not going to pursue further, even though we had gotten up to our target run times and our target performance, because that

21:55 first like month was just to,

22:01 like operationally for them, it didn't

22:04 provide the outcomes they were looking for. And so they'd already decided, and they just kind of let it run out the rest of the fiscal year. So how would, if you had done a TA, how would it have

22:14 said, okay, we've done the Monte Carlo, we look at what the risk is, and if we assume that we needed to fix this thing a lot. You would have that risk covered. And then it would have shown up in

22:25 our budget for the pilot. And unfortunately, we were running too skinny as a business, and so literally, I think we got the thing up and running. It ran for 48 hours continuously. I was like,

22:37 okay, good. We didn't have enough budgets, so everyone flew home, and then literally an hour or two, when we were in the air, the system went down. I had to fly everyone back out the next

22:48 Monday and the reality is we should have recognized that uptime was really valuable to the customer and station kind of on-site support continuously. And we just didn't have the, we didn't bake it

23:01 into our project budget for that first pilot. It's very interesting I. mean, you can do Monte Carlo and reliability is something. You can just, you know, just say, what is a downtime at just a

23:14 parameter and that will calculate the cost with that intermittency baked in, right? Because what's going to happen is your yield is going to be low. Yeah. So it'll get you your levelized cost to a

23:26 certain point. Now you can bake that in, into your proposal for your pilot. And for us specifically, it was not that it would cause our system to have lower costs, it would shut down someone else.

23:36 If our system was down for more than 40 hours, it would cause other things to shut down because of regulatory requirements Exactly. And we didn't.

23:46 I don't think we appreciate it enough, like how important regulatory it is, maybe, is like, that customer has to live by that. And you're bringing out all the complexity that as a founder, you

23:57 had to take all of those things, right? So from how we are offering it, is we are saying, we have thought about all of these things. When we work with you, we'll give you, hey, have you

24:09 thought about this? Have you thought about maintenance? Have you thought about downtime? Have you thought about, you know, the intramatencies that happened in the system and that's what we get to

24:19 simulate, right? And that is the value we kind of add because straight math doesn't work in these complex situations. You can multiply and add and, you know, just simple math because there's so

24:33 many dimensions to it. Too many dimensions, how much electricity is pulled, how much natural gas is pulled. Everything is going to be based on rules and you know, a direct. it's not just exactly

24:47 linear. So you need to understand, and even the smallest of differences, get to affect your cost considerably. So the common example I tend to give is, when you put in a turbine farm somewhere,

25:05 people just assume, here is a power curve, here is it, you multiply it, I'm going to get that power output. Well, you don't, because it's going to be based on the location, the weather, you

25:17 know, how much, you know, there's tremendous intermittency that's baked into wind power. So how do you bring that together so that now you're not affecting a grid, that you're supplying it to?

25:30 And you actually understand what is the output. So because of that, what's your levelized cost of energy? You know, how is it going to affect your returns? But most importantly, how is it going

25:43 to affect the. you know, the transmission because your grid is going to expect a certain amount of power that's coming out of the turbines. So if you look at it, I mean, that level of analysis in

25:57 number of cases has been simplified. And so you go with some raw assumptions that can really affect, you know, the overall, not your overall metrics. So when you go in and say, oh, you know,

26:12 wind power is not working. Well, it's not working because certain, you know, simplified assumptions have been put in place. So how can you get that more granular and you can look at that

26:24 levelized cost of energy? I think more and more companies are doing that, but that's the level of analysis you need to do in order to kind of satisfy your grid operators, your, you know, whoever

26:38 your customers are to be able to say, Hey, I've done this. I've thought through everything. I here is here is how I through everything. And here is the risk that is involved. You know, that

26:51 just gives so much confidence to either your customers or your VCs that you've done your homework. You're not just telling me what I want to hear. You've done your diligence and I can see it in

27:05 metrics. Oh, I get to play with it as well to make sure that I understand and it gives that level of transparency that the industry needs because there's still skepticism in the industry. Are all

27:22 of these technologies going to work? Yeah. And are they going to work in a system, right? Yeah. And I guess what I'm hearing from the story that you shared, Jason, is that some, you know,

27:34 there might be listeners who are thinking, Well, I'm too early for this. Yes. I'm just doing a small pilot that even at a pilot stage and how critical the pilot is, like I realizing if you fail

27:44 at a pilot. you're not gonna go forward to the next stage, right? And realizing that, and maybe you can explain on that, Sujata, like how early can we do a techno-economic assessment? What kind

27:57 of data do you need that you need to have? In clean tech, you know, so let me go back. TEA can be done in three different ways. One is it's using historical data and actually doing the analysis

28:14 and looking at economics. So if you are an established industry with a lot of, you know, prior data, that can be done. Now, if you go into clean tech where there is literally very little data,

28:29 you can get it a lot from public sources, but you have to go with assumptions. And those assumptions come from scientific documents, some, you know, research the innovators or the founders have

28:40 done themselves. And you have to bake that in as assumptions So you now have to bring in. some of the publicly available data with the assumptions and do your TEA, right? There is no escaping this.

28:52 So for clean tech, you actually have to take that second path. And as you mature and get pilots, now you can bring a hybrid thing together, which is getting in the real-time data, looking how

29:06 your rules are projecting the TEA and your assumptions are projecting the TEA, and then you can converge or see where the divergence is, and you can go in and figure out how you're going to tweak

29:19 your system, because your data is saying something after your pilot, while your rules are saying something, where is the divergence and how do I get it close together? Are you ready to lead the

29:31 decarbonization charge? Energy Technologies is your platform for growth, offering unique resources and expertise for energy and carbon tech founders Join us at energytechnexiscom and

29:42 start building your thunderlisset. So I think you see a real split with entrepreneurs, depending on what generation they're in, meaning if they're a first-time founder or a second-time founder.

29:54 And first-time founders usually show up at the technology that's whiz-bang and cool, and they got to find the problem they're gonna solve, 'cause it's new physics or something. You're kind of

30:05 constrained by that. But I think you see with a lot of experienced entrepreneurs who've been in a market and can understand these dynamics, they actually start with the TEA, and they say, okay, I

30:16 really wanna understand a problem in the market, and then we're gonna go build the solution that saved the most money. And a lot of times they will frame out and say, okay, and this is like people

30:27 who look really inter-oppression, they almost do this without explaining it, but they'll look at a changing market, like the good example is the explosion in EV technology. What is that actually

30:37 gonna change? What's the techno-economic analysis gonna look like for other pieces of the infrastructure in three years five years. And so you have these people who look really like they had great

30:48 foresight 'cause they saw, there's gonna be a gap in installing chargers and where's the gap? It's gonna be an infrastructure problem because the grid can't necessarily deploy enough power. They

30:59 knew this years ago 'cause they had done the techno-economic analysis. They understood kind of where the gaps would be where you can create substantial value. And I think people did that either in

31:10 the battery side or the charging side and others because they got deep on saying, where's it gonna be hard for other people to move because the existing economics is this a certain way and there's an

31:23 opportunity to create value and kind of engineer a solution. And if you're three to five years ahead of everyone else, you're gonna have the product kind of available at that right moment when it's

31:32 needed. But you only kind of develop that. I think a lot of the second and third-term entrepreneurs only develop that muscle because they've failed in other examples. Yeah, they're good, yes.

31:43 they're always building their kind of own model. And so I don't want to say like experience entrepreneurs who are looking for the next thing, they're gonna start here. They're gonna start at

31:51 understanding the economics. You start with the TIA actually. Exactly. And then you figure out the product later, right? And the analogy I use is in your everyday life, you use analysis,

32:01 whether you're building, you're buying a house, you're buying a car, or you're looking at investments, you're doing some kind of A TIA, you don't call it a TIA You're looking at, you know, what

32:14 else, you know, goes. And so entrepreneurs should actually start with the TIA. 'Cause then you would also know what kind of data you should be looking for, right? And you might not have the data,

32:26 but you start making assumptions. Exactly. And you start looking at, okay, where are the assumptions? So when I talk to founders, I say it doesn't matter whether you're in pre-seed seed, start

32:37 with the TIA, because that gives you an understanding of how your technology is going to provide. both from a technology perspective, but also from an economics perspective. Importantly, can you

32:49 make money? Can you make money? At the end of the day, that's what this is all about. Can you make money out of it? And make enough money to make it worth everyone's time and effort and yes. And

33:00 with all these regulations coming in, people go, okay, I'm getting so

33:07 much money for direct air capture. So if I build this technology, I can take on this IRA and do this. Do it with an IRA and look at your analysis. Do it without an IRA, look at your analysis and

33:20 see, okay. How long is, these policies are going to be changing with different political environments and it's going to change across geography. So you need to understand how tax all of these

33:35 impact and if nothing existed, what is it going to be? you can take advantage of the regulatory climate, but also kind of look at it from a pricing perspective. And I imagine the hardest thing to

33:48 think about right now for someone who maybe needs to include carbon in their economic model is how the, in some ways the price of carbon today is artificially high, if you have really good high

33:58 quality carbon, and it's unclear what

34:03 the terminal value of carbon will be, meaning in 30 years, Europe has a target, I forget what it is, I'm gonna embarrass myself, if it's like40 a ton, but you see a lot of the news articles

34:14 today, we'll start with like thousands dollars a ton, I think there was a recent one of our guests came on and they were targeting, it was like600 a ton, that's not long-term sustainable, so you

34:26 need to understand, maybe you can sell your cash flow problems in the near term with a

34:34 TEA, but you gotta understand, okay, this is gonna build a sustainable business, And we're gonna have a squeeze in the long term. What do our costs actually need to get to? Can we get there?

34:43 And

34:45 will that make for a good business, right? Which is what we come back to. And I can imagine thinking of that glide path is so hard if you're trying to predict, right? Just what you said, you

34:55 said it's a multivariate analysis, right? It's multivariate, it is multi-year, you know. And so just look at the dimensions you have to think through And so you really do need a comprehensive

35:11 software that is going to kind of guide you through, okay? If you had these tax benefits for the first five years, first 10 years, and then going into 30 years, it becomes

35:22 just regular costs, then how is your NPV going to play out? How should you price it at that state so that now you have, you still are able to break even and make money in your business and make

35:36 money for your customers as well, right? So that is the complexity. I mean, it's really, you know, people ask me the question, oh, can you deploy AI to it? One, you can deploy AI if there's,

35:52 you know, tons of data, which they don't have. So I tell them we are deploying AI. The AI is based on the rules that you embed. So it's almost like deterministic AI that we use in order to build

36:05 that knowledge into your system. So now you can do the TEA Now, all of those don't change, right? When you go from seed to pre-seed

36:19 to seed to whatever alphabet series you're in, this doesn't change because physics and chemistry doesn't change. What you're going to do is tweak the parameters because that's what you're shifting

36:30 along in your journey. Yeah. Do you -

36:35 well, one of the things I always thought would be fun. like a startup nerd is like you do the TEA as like up for a typical customer and a typical project, but one of the things that you really want

36:46 to do when you're selling a customer is build a TEA for them on their asset in that moment. Because it's one thing to have a case study where you're like, oh, there's this hypothetical project,

36:57 but really, you know, if you're sitting down with them and you're talking about like millions of dollars or half a billion dollars, you want to be able to get in there. And it's very impressive if

37:08 you can turn around, you know, do a discovery call with the customer, figure out their parameters, and within a week have a proposal on their desk that says, okay, there's not much money we're

37:16 going to save you in your project. And, you know, it just, it really, it connects the dots for many people, right? So think of it this way, you go into a refinery and say, I'm going to put a

37:30 point source capture technology here. And, you know, you just say okay I know how much flu gas, how much concentration. Here you go. This is how efficient my technology is.

37:42 And this is how much emissions I'm going to save for you. And you can actually do comparative studies there and show my competition is here. And I can show you why I'm here. And that is the

37:55 uniqueness of the technology or the assortment that I use or the membranes that I use that is going to give me the difference. And just using points or capture is the same thing with electrolyzers or

38:07 it is the same thing with direct air capture. You really need to show the uniqueness and not just be like ununique, but this is why I'm unique. Not only is my technology efficient, but I'm going

38:21 to save you money.

38:24 And save you money not only with tax credits, but in the power you're consuming, or the natural gas you're consuming, or whatever the parameters, whatever it's going to be. This is how I'm going

38:37 to save you money. And that is the storytelling that you talked about. So there's a techno part to it and there's the economic part to it, right? Can you talk about like how is that integrated?

38:50 'Cause you talked about like being able to prove to your investors, for example, it's going to be more efficient, the technology is gonna work. And then there's the economic part of it, which is

39:00 like the numbers are gonna work. It's gonna make sense as a business model. So how do you integrate the two? So the technology part of it comes from, okay, my technology has so much capacity, it

39:15 can absorb, you know, so much input into it. Here is the efficiency of it. It runs so many cycles, it consumes so much power, it consumes so much gas. Or it's, you know, it's creating a

39:31 simulation of the technology that works. I think it just makes a technical sense like you're really parameterizing.

39:38 It's not like the competitor or the other, the many competitors in some ways. And it's not just as simple as,

39:48 like the result may be LCOE, right? A lot less cost of energy. Exactly. And you're targeting whatever it is, 10 cents a kilowatt hour, five cents a kilowatt hour. Like that hides all the detail

39:59 of like how you get there. And like as much as we talk about the grid costing, whatever it is, six cents. That's really an amalgam of solar and wind and peaking power and base load power. And it

40:14 just kind of happens in our current system that it ends up being about six cents a kilowatt hour. But that's like the common, I guess criticism of our current PUC system at ERCOT is like we don't

40:27 value dispatch as much as we should or reserve power. And so battery is just kind of like don't have a life. or a way to monetize it as easy, unless they're behind the meter. And that's why you

40:41 get these kind of weird instabilities in the spot price of the grid because we haven't necessarily reserved. You don't get a base load there. Yes, and we don't value the fact that base load might

40:51 be stupidly cheap to produce if it's always running. Yes. But like that reactor power actually has value in the system. And so, I don't know where I'm going with this. I think that's what makes

41:01 it tech to know economic is you're going deep and saying this is what's driving You can go into it and say, here's a simulation. You can put it behind the meter. You can create base load and it's

41:11 this much. Or you can just bring it. You can have different technical solutions that can give you different costs. It's not about just that endpoint number. It's about like all the things.

41:18 Bringing it all together, yes. And you

41:24 think about the power today and you talked about battery. So it brings, I'm talking to founders who are trying to create either you CO2

41:34 technologies to actually say, I can create arbitrage opportunities, you know, for companies. Yeah. You know, companies can use their CO2 to actually create arbitrage opportunities. So deploy

41:47 the power when the grid needs more power, but you really need to run the full analysis to say,

41:57 how much power, how much storage do I have? And, you know, what is the price of power? And so when do I deploy and when do I use the grid? So all of this is really a lot of parameters coming

42:10 together, data coming together to actually just say, hey, here's a recommendation. I mean, the recommendation might sound like simple, but the amount of work that goes in to actually bring the

42:21 technologies together, superimpose economics on it, and then give you a recommendation saying, hey, during these hours go and sell, during these hours, bye. And that is going to create a good

42:35 return for you at the end of the day. As you said, businesses are in companies are in to make money. And if they don't make money, even in the medium term, short to medium term, clean tech is

42:51 not going to take off. No one's going to

42:54 invest in clean tech for the sake of clean tech, even Europe, where, you know, there is so much push for all of this happening It still needs to sustain. It needs to make economic sense. Yes.

43:07 That is fundamentally front and center to everything. And that's why, you know, we thought of it. I mean, we didn't just say, oh, I'm going to build a cider to do a TDA. We looked at, oh,

43:19 clean tech is hard. It's, you know, it's mostly first of a kind. And, you know, it needs to make sense for customers to adopt. So how can we make it easier for everybody in the ecosystem?

43:32 Founders to. investors to

43:37 companies adopting this technology, how do we make this season? So because of the complexity. So one of the things I wished I had, I guess when we were going out and discovering kind of the

43:49 markets we wanna get into, was I wanted a quick way to build an economic, a TEA essentially for applications. 'Cause I remember we were sitting in front of a whiteboard, it was kind of like a blue

44:02 sky exercise of, where's Power Expensive, where's Reliability a Challenge, and who has money, like the things that we did, and we really just did a force rank matrix, that was, we think it's

44:15 expensive here, we think these guys have money, but we never actually delved in and said, okay, what's the application space where, This makes no sense. We can define the killer app, where it

44:26 makes the most sense, because it was just too expensive, like we had to learn, we had to like compile too much desperate information.

44:33 good control of the technology, we knew where the costs were, and it would kind of like play out. And

44:40 it was, we ultimately had to go in deep on like one specific application, 'cause there just wasn't enough capacity to explore. All the different variations. Yeah, the different dozen things we

44:53 had come up with, and it was really like a swag from the gut of like, who do we think has the most cash today? And I'm not convinced,

45:05 you know, it probably took us another 18 months to get deep enough to know like what the parameters of the product would be. I'm just thinking to myself, if we had,

45:17 like we could have saved so much time, right? That's really what it comes down to. It's efficiency, right? Yeah. Do what you do well, and you know, then you're going to have these supporting

45:27 technologies that is going to kind of help you get further with what you do. you do. And just think of it. If you're able to think about it and say, if I had so much gas and how much power am I

45:41 going to do? And if I go to a different location, if I had different, you know,

45:49 kind of metric of gas coming in, cubic meters of gas coming in, this much power. And you can just kind of look at it and say, I'm best suited based on the metrics for the customer And for me,

46:02 this is my best, this is the amount of optimal amount of gas I need to get to be profitable for myself and be the most profitable for my customers. Yeah, yeah, yeah. And that is what we do,

46:16 right? We, you know, there's a, we did an energy planning for a customer to say, okay, there is so much, you know, they wanted to say, okay, our technology is primarily running on electric

46:29 power and so we can pull from the grid which is going to contribute to emissions or we can use solar and battery. So what is the best thing to do here? So we, you know, we did a report. They said,

46:41 okay, here are the locations. And we went in and said, okay, for this location, do this. And we even were able to recommend this is the amount of solar power you need to, you know, put in

46:53 because anything more is going to erode your, you know, LCOE without getting any benefits. So we did this a little bit when I was getting solar in my house. So there's this like weird challenge

47:07 that we, you know, you decouple power or retail from generation. And Texas, I think famously has no like real net metering. And as a result, if you produce extra access power, you actually

47:21 don't get it back. And so when we were looking at installing the solar, it became very clear to me that you don't want to like buy too much solar panels for your house because you'll overproduce,

47:36 and it doesn't contribute to - Yeah, it doesn't contribute to anybody. And do you not make metric for you? Yeah, for me, it just doesn't pay off 'cause we get no credit back. Yes. And if I fill

47:46 out my batteries, the batteries are full. That doesn't even do anything for us. So we're just wasting power. So there's no reason to oversize. So then the question was, what do you undersize for?

47:58 If you were trying to do it for resilience, you would oversize your panels so that they would produce more power than you need on the darkest day. That would be the engineering way to cover yourself.

48:10 But in this perverse kind of system where we're attached to the grid, the grid's always there. And we don't get anything, any benefit of feeding power back in. So you almost want to put in the

48:20 smallest solar panels as you can. And to get your emissions to a certain level. So one of the recommendations - Well, we don't get paid for emissions as homeowners. Yes. So there's no rec, right?

48:29 Well, the way we tell comes. companies are when you're doing this, we compare it to the price of grid, right? When your price is very low, you know, you would put in solar just for emission

48:43 sake, right? And what happens there is you can buy recs to actually reduce it, which might increase your topics, but you don't have to put upfront cash to put your solar panels. But when your

48:58 price is high, then actually investing in solar panels for the right capacity, optimal capacity, is actually going to bring your LCOE down.

49:11 So this is the kind of analysis we get to do literally in Mad roof an hour. So we're always going to end with this in terms of our house cases. We ended up putting about 50 of the solar we needed

49:27 because really the major benefit was getting the batteries and it was an ITC component.

49:33 And once we did all the math on the LC, we had actually ended up being a little bit more expensive than grid,

49:39 but we're not wasting any power. But what we got out of it was, it was essentially the resilience of the backup battery. It was like free storage, or if we ever lost power. So we almost got the

49:51 resilience for free. And that's where I convinced myself in my head, 'cause this is better than buying a generator. Exactly, that's

49:58 what most people are doing now. If I put a solar power. No, like in this situation, it was like, put it in a small solar and go big on the batteries. And then we sized the solar actually just to

50:05 run the refrigerator and top off the batteries during the day. So you didn't need to run the whole house. We were just sizing it for the situation if we lost power for like two days, right? So it

50:16 was a very different techno economic analysis, totally not what the sales rep was trying to sell me on. He was like, you're gonna produce power and it's gonna go up every year. And when I came

50:25 back with a smaller quote, he was not happy because I did the analysis, but even then it was a swag We didn't do any depth of it. assessment and I had all the historical data on our consumption but

50:37 I didn't have any like like sun data so I didn't even have a really good way to estimate it but it's like one of these what I'm getting at is like if you have that techno economic analysis you kind of

50:48 get these results which are counterintuitive as long as you know what you're optimizing for and what you know what the actual benefits are and you can make better decisions so as we're coming closer

50:56 to the end of our time because this was um you know we could keep going on this actually um practically as a startup um you know how how do we work with Decider you know it's a software as a service

51:11 yes um you buy it as a subscription like you buy it as a subscription and um we make it easy so you know we said so if we have to live with what I just said that different stages of companies earlier

51:25 you start the better so the way we're working with this is the earlier the stage of of the company, we kind of try to give a break on costs. I'm a fellow founder. I know how hard it is for founders

51:38 at the earlier stage, but since techno-economic analysis is so important, we tend to work with founders in the earlier stage by giving them the break on subscription and it increases as they go into

51:52 the different stages. The idea being, I'd love to be on their journey as they go in and get the technology to scale And

52:04 the way we think about it is a lot of companies, whether you're a C-stage company or you are a series B company, they tend to use consultants to come in and do the analysis. Great. And what I say

52:20 is,

52:23 make sure that they are delivering it in a sustainable technology like Decider because what you're going to be left with is a whole bunch of.

52:34 PowerPoint with an Excel that is left behind, that is great for that time, but you're never going to be able to use it going forward. So how can, you know, so this is really about creating the

52:50 scale and creating that sustainability of your analysis moving along with you in your journey. So just think of it as you're working with different players. That's an advice. I mean, I can always

53:04 talk about using Decider, but it's like, you almost want to invest in infrastructure for the business, not a report. Yeah,

53:12 and that's why having a robust tool can be, you know, valuable along the way. Absolutely. So one of the key things that it just came up and I really wasn't thinking about it when I built Decider

53:25 is how startups can can actually use this technology for their BT or sales, right? So you talked about it, right? Just imagine you have this analysis, you just log on, and the customer says, oh,

53:40 I have this configuration, this configuration, they are able to just, within minutes, just put it in, and they say, okay, here, this is how it's going to look. In a very visual, with all the

53:53 metrics, I mean, that is a powerful BD tool. Yes.

54:01 Sorry, guys, Google decided to pick up and ask questions. I don't know whether it's my, it's your series. It's your series. It's your series. Siri cannot figure out technology analysis. We

54:14 have to train Siri. But, you know, from BD to creating proposals, you know, efficiencies, the name of the game. And once you do it, everything else becomes so much more efficient. And the key

54:28 also, in some ways, when you're doing these complex sales, it's not just your champion you have to give this to, they need to afford it to their boss, you gotta forward it to the FID committee or

54:37 group. And so the more portable and what do you call auditable, maybe is the right word, or like being able to kind of dig in and see what's really working. Like that helps convince the right

54:49 stakeholders that this is real. And so for that particular reason, we created workflows. So you can just share these, we created automated reports So you can do the analysis and say, print PDF,

55:03 and it'll create a report that you can easily send over if you don't want to send the direct analysis over, right? So we talked to so many different customers across the ecosystem and tried to

55:16 create all these efficiencies that is going to make it easier for better adoption and, you know, creating a cleaner world is not easy So, how do you. kind of help with different technologies that

55:34 can kind of lead you there. Yeah, we need to grease the skids on the capital of appointment. Yes, exactly. And can you talk a little bit about the onboarding? Do you have someone walk through

55:46 them on how do you use Decider and over the journey of it? You know, how independent do you get? Do you need a consultant along with it? So for Decider, what we said was we have a team in India

56:01 of PhDs in chemical engineering and physicists who actually kind of work with us. So when a founder comes to us and says, okay, we are doing this kind of technology. We want to create, this is an

56:16 actual truth story, which is they said, okay, we want to be able to create

56:23 methanol from this feedstock. And so we actually ran the full analysis team in India will actually create the rules. This is the feedstock. This is the different process that goes through. This is

56:36 the output you're going to get. This is the emissions. All those rules are actually written in a report and given to my data scientists who then embed those rules and provide it back. So we work,

56:48 we want to make it easier on founders. So the onboarding is the founders say, this is the technology. This is what we want to see. And then we take the rest And then we actually provide citation

56:60 on where all these numbers are coming from, which is also a lot of transparency. And so when you talk about putting this in front of investors, it's almost like having a third party. I don't know

57:09 about a certification, but there's like a third party who's put in

57:15 reasonable numbers. Yes, better numbers. I mean, I cannot say we are 100 accurate because we are relying on scientific numbers and research that we come up with but what we are doing is we are

57:26 doing the heavy lifting.

57:30 you know, our technology has gotten so far ahead now. A lot of the problems that the customers come up with, we've already solved it. So in which case, we don't even have to do the research, we

57:42 just deploy it. But for new ones that come in, we actually do the research and we kind of work with the founders and say here it is. The biggest heavy lifting for founders is validation.

57:57 Exactly I mean, you can't bypass that. You know, the founders have to be comfortable, they can walk that number and we need to prove to them that we can walk those numbers with them. That is the

58:10 biggest heavy lifting that founders have to do, which I think is reasonable because they have to be, it's their technology, they need to be comfortable. So how do we sign up?

58:21 Absolutely, you know, just,

58:26 you know, reach out to me, right?

58:32 Sujatadeciderapp, go to my website, get on my LinkedIn post, we are constantly posting talk leadership. Our journey is the journey with every heart tech founder in clean energy or clean tech, I

58:49 should say. That, I believe, is going to be a very gratifying journey for me. You know, for me, I have to connect the ecosystem, but the founders are close to my heart.

59:04 Thank you so much for sharing. Hey, thanks for having me on. And, you know, having this kind of role play is actually very nice. It's great. I thought it worked really well. Yes. Good. Got

59:15 to be here and contribute.