#123 Driving Adoption in Hard-to-Abate Sectors, Why Not an IP Play, Fundraising for an R&D Shop, Partnering w/ Large Corporates, Hiring PhDs, & More w/ Jeff Erhardt (Mattiq)

In this episode, Silas Mahner (@silasmahner) & Somil Aggarwal (@somil_agg) speak with Jeff Erhardt from Mattiq.

Jeff had a ton of insights and we are glad to share them with you today.

Enjoy the Episode! 🌎

📆 Book a time w/ Silas during Climate Week in NYC (IRL): https://calendar.app.google/STXX7pMNmsWRsafd6

📺 👀 Prefer to watch: subscribe on YouTube.

📫 Interested in written summaries and takeaways from the episode? Subscribe to the newsletter.

Want to be part of the community and engage further? Check out the Slack Channel. https://tinyurl.com/mwkn8zk5

-----
Links:
**Connect w/ Jeff: https://www.linkedin.com/in/jefferhardt/
**Follow Mattiq: https://mattiq.com/
**Climate Capital Summit: https://www.climatecapitalsummit.com/ 
**Check out our Sponsor, NextWave Partners: https://www.next-wavepartners.com/
**Follow CleanTechies on LinkedIn: https://www.linkedin.com/company/clean-techies/
**HMU on Twitter: @silasmahner

-----
Other episodes you might enjoy:
**Most Recent Episode: #122 Offset Skepticism, Regulation Driving Demand, Sustainable Supply Chains, Content GTM Strategy, Climate Business Opportunities, & More w/ Saif Hameed (Altruistiq)
**Similar Topic: #113 Incubating Climate Risk / Insurance Solutions as Traditional Insurers Pull Back w/ Stephen Brittain (Insurtech Gateway)
**Something Totally Different: #99 Capturing Methane to Power Data Centers in Remote Areas w/ Andreas Lehner (PowerBlocks)

Chapters

• 0:00 Mass Commercialization of Sustainable Materials Engineering
• 5:25 Industrial AI
• 19:00 Clean Chemical and Fuel Production Development
• 26:58 Using Electrochemistry for Clean Chemical Reactions
• 36:23 The Challenges of Research and Commercialization
• 45:46 Business Models and Partnerships for Sustainability
• 53:02 Opportunities in Clean Technology Industry

Transcript

Jeff Erhardt: 0:00

And I think one of the things that I've both seen and heard from a number of different companies that raised money over the last, you know, say, four or five years working in this space, thinking about tackling sort of the clean chemical fuels, is they didn't go that deeply down the economic analysis.

Silas Mähner: 0:23

Welcome back to the Clean Techies podcast, where we interview climate tech founders in VCs to discuss all things building and investing to solve the biggest challenge of our generation climate change. Today, somail and I had the chance to speak with Matique, or also called Matt IQ CEO Jeff Earnhardt and how they are mass commercializing R&D for sustainable materials engineering. They do this by first looking at the areas in need of decarbonization, applying basic principles of physics to know which areas are most feasible to develop new technologies in, at least currently. Then they use AI to get in the neighborhood of possibilities through simulations and once they are in the neighborhood, they have a methodology of rapidly testing the technology in the lab. To go through the R&D phase a lot faster than most places would, thus reducing the total time to get these technologies off the ground. It's a really beautiful example of how powerful a team of highly technical experts can be when you pair it with a solid commercial team and a good partner. Outside of what they do and how they work, Jeff also shared advice for early career professionals, the misalignment of incentives in academia to go for the moonshots, why they chose a partner with companies to develop technology rather than just sell the IP. Some ideas of areas to build in climate and the importance of identifying and hiring, as he calls it, icons that will bring their followers with them, especially when you're hiring technical talent like PhDs.

Silas Mähner: 1:50

Generally speaking, a great episode. Let us know what you thought once you finished listening and enjoy the show. All right, Welcome. Welcome to the show, jeff, good to have you on Thanks.

Jeff Erhardt: 2:00

Thanks for having me.

Silas Mähner: 2:02

It's a pleasure. Where are you? Where are you calling from?

Jeff Erhardt: 2:05

I'm currently in Chicago.

Silas Mähner: 2:07

Okay, in Chicago. I'm originally from Northern Wisconsin, so not too far.

Jeff Erhardt: 2:12

You know what summers like around here.

Silas Mähner: 2:14

Yes, it usually comes on a Wednesday and then it's gone.

Somil Aggarwal: 2:18

Exactly.

Silas Mähner: 2:20

Very good. Well, I'm glad to have you on, so let's dive right into things. Give us a quick background on who you are and what you're doing and how you ended up at this point.

Jeff Erhardt: 2:29

Yeah, sure, absolutely so. Yeah, so my name is Jeff Harhard. I'm the CEO of Matique. We're in the process of building what I like to think of as the world's most technologically advanced clean chemistry company. The background in our company is we were spun off from Northwestern University, the Department of Chemistry and Material Science there, where they've been doing some very interesting research on fundamental materials and so helped to spin that out about two years ago. I joined in early 2021 to help build and commercialize and lead this, and so we're really on a mission to drive this company to work on decarbonizing the production of chemicals and fuels which are responsible for a pretty substantial high single digit almost 10% of global emissions. So that's what we're up to.

Jeff Erhardt: 3:13

My personal background I was originally a material science by training, but I was not a chemist. I was a solid-stake physics guy. I started my career in the semiconductor industry doing R&D for them back before semiconductors got cool again. So it's nice to see that renaissance. But I started off doing that, and what I got stuck into doing was building large cell software systems to help drive the R&D production of a couple of billion computer chips every year, and so that sort of led to my transition from being a pure visual sciences guy to being a software guy. I then took some detours into the startup space, built a few companies in the enterprise software space around platform analytics, ai, machine learning and took some time off before joining and helping this team. So excited to be at it.

Silas Mähner: 4:04

So I guess I am kind of curious. Obviously, matik has a really particular focus on reengineering materials, so I am curious have you always been interested in the climate space? Is this something like? Oh man, this is now the opportunity for me to do this. How did you get interested in the climate problem at all?

Jeff Erhardt: 4:21

Yeah, I think the I could answer in a couple of ways, you know. I think the first one is right, wrong or different. I've always been drawn to what people might call hard tech or deep tech type of problems, really challenging problems that are sort of meaningful to the world and gives me the chance to bridge deeply technical, scientific people with challenges that need to be solved in the marketplace in the world. And so, as I was taking sort of my time off, I took some sabbatical before doing this. You know, sort of over COVID I had an hour or a year, year and a half of time to go, think, to study some different problems, sort of this broad challenge of energy transition. What does that mean? Sustainability started to become very interesting and I didn't know what exactly application I wanted to go after, but I knew that for this next, next endeavor, I wanted to be something meaningful and especially seeing, you know, the challenges arise out of COVID and then the increasing challenges associated with sustainability, climate change the two really came together nicely.

Silas Mähner: 5:25

Hey there, quick break to remind any founders or VCs listening. If you are looking for deal flow, seeking to raise funding, looking for partners to help service your needs, or perhaps you're looking for corporate investment partners, feel free to reach out to us through our Slack channel, which can be found in the description. Because we meet a lot of people in this space, we set aside time each week to make introductions to the various people that we encounter. This is something we do free of charge in order to help these incredible companies solving climate change to scale. Looking forward to hearing from you in the Slack channel.

Somil Aggarwal: 5:54

Would you say, for your interest in deep tech. Was it a natural transition or was it something that you found to be more intentional?

Jeff Erhardt: 6:02

Yeah. So I think I guess, whether you could say it's intentional or a natural transition, I've really been doing that fundamental in my whole career. So if you were to put a hat that I were to wear, is I'm not fundamental? I refer to myself as a recovering engineer, so I'm not nearly as smart or as technical as all of the people that I work with these days, but really what I have the ability to do is to bridge the gap between people who are doing deep scientific research and development and then figure out how to craft that into a market need that solves the major problem, and so for me that's kind of been a common thread throughout. What I've done is creating this bridge between deeply technical people and market need and your background.

Somil Aggarwal: 6:44

One thing that's really interesting about it is you had a position at GE that was related to industrial AI and it seemed like it followed the acquisition of your startup, Wiseio. So for the audience, I would love to hear the background of that company and also what is industrial AI?

Jeff Erhardt: 7:00

Yeah, sure, absolutely. So that was a really fun company, and it's a lot like this. One of the things or one of the things that I like is history may not repeat itself, but it certainly rhymes, and that's certainly the case here. So the history of Wise was originally incubated at UC Berkeley. This was in the sort of early 2010s, and it was led by a professor there, his research group, by the name of Joshua Broome, who is an astrophysicist, and the problem that he faced was they were getting all of these telescope images of the sky, massive amounts of data, and they couldn't hire enough grad students to be able to sift through them and find very subtle patterns in this data. And so what they did was they started to train computers to do that, and so the very simple way that I like to think about machine learning or AI is the process of teaching computers to mimic human decision making on top of data. And so, anyway, with that company, what we did? We spun it up from Berkeley and we started to build it up into one of really the first wave of machine learning called applications companies.

Jeff Erhardt: 8:15

Today, it seems obvious Everybody is doing AI for X, especially with the new wave of generative AI and chat, gpt, etc. But at the time, the concept of embedding machine learning or AI, as we call today into specific applications was fairly novel and fairly fairly rare. So we had to do a lot of work on building the infrastructure, the methodologies, the tools, that templates to make that happen. Anyway, long story short, we built that up. It was pretty successful, and we started to attract the attention of a number of different companies. The one I didn't expect to reach out to us, as people were interested in acquiring strong machine learning companies, was an old line industrial company by the name GE, and this was around 2015, 2016. And they were starting to go through their transition of thinking about how do we create a digital industrial company, ie, how do we take our deep expertise in these things in the physical world jet engines, power turbines, mri machines, windmills for renewable power generation and how do we start to make those processes, that equipment and the things that they do better by applying software to it. So software data and then eventually, once you generate tons of data using machine learning and AI to make sense of that. So, anyway, we joined that company in 2016, and then helped to build that practice, exactly as you said, which is this concept of industrial AI, then the natural question might be well, why is that different than applying machine learning or AI in other fields?

Jeff Erhardt: 10:05

I think there's a couple of things that are really important to understand, because, even with all of the interest and excitement that's happened over the last, say, six months with the generative AI in the large language model, it's really critical to understand where those are useful and where they excel, and what some of the challenges are in applying them in other markets and other solutions. Though, in particular, the way that I like to think about these, as I said earlier, at the most basic level, is teaching the computers to mimic human decision-making on top of data. This is really accomplished by doing this, called training these models to identify patterns in large amounts of data and then really making a best guess this is the key point an imperfect guess as how to answer a question that is asked of that model based on the patterns that were previously identified. So you can think about the way that works in these current so-called large language or generative models that are coming out. What do they do? They train their models based on all the publicly available information on the web, and then they're using them to interact with people and understand and ask questions, much like you would go and use Google search. That's good, that's useful, that's interesting, it's very powerful and people are using them for lots of good things.

Jeff Erhardt: 11:27

But the industrial world is different for a couple of reasons. The first is that, simply, the data doesn't exist in the connected, available and aggregated way like it does on the web. Now that's certainly starting to change and there's people taking great efforts to connect their equipment, to aggregate this together, to make sense out of it, but that is a real problem. That needs to be considered is the fragmented and imperfect nature of the data that exists in the industrial world or in the physical world. The second thing that I would say is that the physical world is just that, is, that it's, by its very nature, based upon the fundamental laws of physics, as well as bounded by them. So when you think about applying data and math, ie AI to that, you can't ignore the behavior or the fundamental physics and you need to think about the interplay or the interoperability between the physical world and what makes sense in the real world and what is predicted in a model which is fundamentally an imperfect approximation of the real world. Then, finally, the last thing that I would say is you really have to consider what is the cost of being wronged and what do I mean by that?

Jeff Erhardt: 12:51

So I mentioned that when you're asking a model a question, it's simply making a prediction, it's making a best guess as to what the right answer is.

Jeff Erhardt: 13:01

That's got some uncertainty around it In certain spaces, for example, if you're serving an ad, if you go to a website and they serve you an ad, it's not something you're interested in.

Jeff Erhardt: 13:13

The cost of being wrong isn't very big, it doesn't really matter.

Jeff Erhardt: 13:17

But in other things, if you're thinking about things in the healthcare space or things in the industrial world, one of the problems I spent a lot of time on, which is how do you do maintenance on an aircraft engine, and deciding whether to take that aircraft engine off the wing and do some maintenance or not is a whole lot different if you get that prediction wrong than it is if you serve somebody the wrong ad.

Jeff Erhardt: 13:48

So, really, the combination of those three things together the fact that the data is simply very fragmented, often dirty, imperfect and not very well connected, the fact that these processes, by definition, have to be grounded in the fundamental laws of physics and what proper behavior is. And then, finally, the fact that making imperfect predictions could have very dire consequences really means that this concept of industrial AI can be based on many of the same principles but needs to be considered and treated in a very different way. And that really leads a lot of what we're thinking now about at the company and how we can leverage some of those thought processes, those capabilities in this interplay between the physical world and the digital world to solve some very big and new challenging problems that the world is facing from a climate change sustainability standpoint.

Silas Mähner: 14:45

Hey, there are you building a climate tech business and looking for very specialized talent? Consider reaching out to our sponsors, next Wave Partners. Next Wave are experts in talent acquisition, recruitment and retention across the climate tech, renewables and ESG spaces globally. So if your team is growing or you're looking to make a career change yourself, feel free to reach out to Next Wave at Next-WavePartnerscom, or reach out to one of their consultants directly via their LinkedIn page.

Somil Aggarwal: 15:08

That's really, really cool because I think it's going to be really interesting to see how that plays into the way that you think about the problems that you're solving now. I think the point that you brought up about this being industrial context being more critical, or at least where AI and data can come in, is, generally speaking, in critical processes. I find that really interesting because that's the barrier that we see to a lot of traditional investing within climate, which is, in order to disrupt that traditional barrier that exists, you probably need a lot more investment, but at the same time, it's chicken and the egg, because how do you get that investment if you're projecting a longer time scale? You've had now, obviously, a current success story, a previous success story operating in this. What would you say worked for you in terms of being able to be the person that convinced a hard line sort of backer or an industry that was lagging or required more convincing to adopt your technology? What do you think really drove you to the finish line?

Jeff Erhardt: 16:07

Yeah, I would say really three things, which is patience, persistence and pragmatism. And so what do I really need by that? There is a reason why traditional industries those industries I worked with when I was at GE, and those industries which still need to go through this transition of becoming more sustainable, just like you were asking still do things a certain way, and that's because they have tremendous knowledge, expertise, about what can go wrong and what are the consequences of something going wrong. And so, really, what I learned is, again those three things right, that it takes time. And so what does that mean from a startup standpoint? That means how do you set realistic expectations, how do you manage your burn rate to be able to recognize and realize that transition and change in the physical world, especially in the traditional industrial world, moves slower than it does in the digital world and sort of modern digital first industries?

Jeff Erhardt: 17:14

Persistence means that it's going to take a while to get things right, and there is iterative experimentation that's needed and you need to think about how do you crawl, walk, run in terms of what the solutions are.

Jeff Erhardt: 17:29

And then pragmatism means a couple of things. Pragmatism means what does it mean to work in the constraints of the way the world is run today by these legacy companies and, in particular, one of the things that we spent a lot of time thinking about and optimizing for is that, well, it's good to drive these sustainability solutions for the sake of the world and for the sake of sustainability. If they aren't able to stand on their own two feet, ie if they can't be economically viable or commercially competitive on their own, it's very hard to get people to adopt them, and so one of our key monitors as we're working on developing these solutions for producing clean chemicals and fuels is that they must be economically viable, and our goal is to eliminate the green premium that so many people have made the bet on in these sustainable solutions that are being promoted today Interesting, so I want to follow up with this.

Silas Mähner: 18:34

I think Thanks, just to maybe reiterate I don't necessarily know if we need more clarification because it'd be getting very, very technical and nuanced to the particular situation. But to reiterate the point being if you think about your potential technology and you think about the constraints of physics and you take it out to the world, can you actually make it financially feasible? If you cannot, probably try something else? Right Is basically what you're saying.

Jeff Erhardt: 19:02

That's exactly right, and what we are doing is we are explicitly using our technology, our differentiated technology, to develop these type of electrochemical systems, ie, clean production of chemicals and fuels to eliminate that green premium, solve the ones that are most economically viable, generate more expertise that we can build upon, leverage things like data and AI to understand the correlations between those reactions and then, over time, start to address those ones that are harder and less economically feasible today.

Silas Mähner: 19:39

Yeah, that makes sense. So basically focusing on where the most value can be had as early as possible, because we don't exactly have a lot of time, that's quite interesting. I'm sure it'd be great if you have any resources we can drop in the show notes for people afterwards.

Jeff Erhardt: 19:53

Sure, no problem.

Silas Mähner: 19:54

There's a lot of companies building in this space and they probably could take a step back and look at what they're doing. So one thing I do want to get to before we move on to more of the company specifics is you've had a pretty interesting career and worked in pretty monumental, really interesting spaces. For that. Do you have any thoughts looking back to offer early career professionals right now on how to consider what field they're choosing to work in, based off of your experience, let's say, choosing right. Whether that was luck or intention, I don't know, but let's hear your thoughts.

Jeff Erhardt: 20:28

Yeah, it's probably more like the former than the latter, so, as I think about that, it's both. Again, you can pick your Yogi Berra quote or whoever the quote is about success being the combination of luck and preparedness. But in my case, as I spent a lot of time doing this mentoring, talking to giving talks for early career or in college people and a bunch of different themes come out, but can highlight a few here. The first thing I would say is don't assume that your career trajectory is going to be linear or predictable, especially over the course of a 20 or 30 or 40 year career. What is hot today, or what you're good at, or what you enjoy today, is very highly unlikely to be what's going to be hot or what you enjoy before. And so, just to give an example of that, I started out my career as a material scientist doing semiconductors and by the mid 2000s people said that industry was dead, it wasn't interesting. So I switched and I started building and running software startups and that was fun and that was interesting. But now you start to look backwards and what did two things happen. First of all, that's the hottest industry now in the world. If you look at NVIDIA's earnings and people wishing that would come back to domestic production within the US. But, more importantly, I was then able to bridge and bring together both that physical sciences, material science background as well as the software development, machine learning, ai background, and so my only point on this is don't assume that exactly things are going to be predictable or that it's going to be linear, and you should embrace that.

Jeff Erhardt: 22:18

The second thing I would say is to really look for analogies. It was sort of said before history may not repeat itself, but it certainly rhymes, and one of the things I was taught early on in one of my first startups. I was spent a lot of time working in the financial services industry, selling to investment banks, hedge funds et cetera, and one of the people there told me is who we're talking about, who is hiring, what talent he was looking for and why they were interested in using our software. And his point was I can teach people finance, but I want people who can think in other industries outside of mine. So he wanted to be able to hire biology people, history people, et cetera that could think and can learn the finance and the math, and the point was if they can pattern match from those other fields, it can give me a unique competitive advantage. So that ability to recognize ah, I see how this works in the field of aviation and now I can apply it to chemicals and fuels production, for example is very useful.

Jeff Erhardt: 23:31

And the last thing I would say is really, especially early on, find a great mentor, find somebody who is on an upward trajectory and think about making them successful. And I think it's really powerful if you can change your mindset into not just thinking about your own success but thinking about somebody else's success, and then that has a way of providing a slipstream and helping pull you through a career. So again, just to think about all of those things in general, right, assume there's going to be twists and turns and not everything's going to be predictable. Think about what are the analogies for how things might play out in the future based on what has happened in the past, and then think about really skating to where the puck is going to be. And then, finally, you know, try to change the mindset from only thinking about your personal success to thinking about the success of somebody who cares about you and seeing the goodness that will come out of that.

Silas Mähner: 24:32

Hey there, thanks for listening to this episode. If you made it this far, it's likely that you're enjoying the show, so I wanted to ask your help. If you're enjoying it, please give us a review on Apple podcasts and share with somebody in the same industry who might find this interesting. And if you're interested in getting summaries of these episodes, go subscribe to our newsletter that comes out on LinkedIn, and Substack Links can be found in the description. Thanks for your help in growing the reach of this show.

Somil Aggarwal: 24:54

So, yeah, I really appreciate breaking it down and sort of fitting in the variability that will inevitably happen and the ways you can kind of hedge against it. I know that that's oftentimes what feels most applicable is, you may not be able to control the luck, control the chance, but try and manufacture it right and try and associate with the people that can help you get there. So I appreciate the candor there. You know, I think we've set the scene very well in terms of understanding who you are, where you come from and as well as some of the previous successes that you've been able to have. I would love to now transition to just, you know, fully open the, open the FLO gates and really talk about a mad IQ.

Jeff Erhardt: 25:30

Sure, absolutely.

Somil Aggarwal: 25:32

Is that? Am I pronouncing that correctly?

Jeff Erhardt: 25:34

You can say it however you like. I'd say. I'd say I'm a TIC, but certainly the IQ infers that there's some intelligence under the hood and are studying of some of these critical materials. So either one is perfectly fine, I'm not offended. Well, yeah.

Somil Aggarwal: 25:50

Go ahead, Go ahead. No, no, please Well. So I appreciate the clarification there. So so, Matique, could you please talk us just through what you know, at a high level, Matique does and different the different areas that it covers? Yeah, sure, Absolutely.

Jeff Erhardt: 26:05

So, you know, I think, as I mentioned when we we picked off, you know we're really on a mission to build what I like to think of as the world's most technologically advanced clean chemistry company. And specifically, so what does that mean? Right, you know, we really envision a future where the world transitions from producing these critical chemicals and fuels that power our world, where they're manufactured using fossil fuel feedstocks and fossil fuel heat, into a world that leverages renewable feedstocks and renewable energy, and, in turn, you know, that's going to really allow us to decarbonize these, what are referred to as hard to abate sectors that account for close to 10% of global emissions. So how are we doing that? You know we're doing that by really radically transforming the world of electrochemistry. And so what's electrochemistry?

Jeff Erhardt: 26:58

Electrochemistry is starting to get hinted at earlier. It's the idea of using electricity to drive or to power chemical reactions instead of heat, and, on the one hand, it's both a technology that's been around for a long time, has quite a bit of history, but it's also a very new technology, and so the reason it's getting a lot of attention recently and more mainstream attention is because it's being used in the production of clean, or what's called green hydrogen in a process that's called water splitting. So the basic idea is you take water, you put it in one of these electro catalytic systems. I like to say you plug it in the wall, right, you give it electricity in the context of a catalyst. In the case of water splitting, that catalyst is a very rare element called a rydium, and then two things come out the other side you get oxygen, which of course isn't harmless, and you get hydrogen, which can be used as a clean fuel. And so really our goal and what we have built is the ability to study these very complex systems where many, many different elements and variables need to come together from the specific catalyst that is used and the materials that define that catalyst. What is the membrane, what's the electrolyte, what's the operating conditions, what is the chemical reaction you're going after?

Jeff Erhardt: 28:24

To date, those processes and those systems have only been studied in a very linear, ie one at a time, in a very fragmented way, by many different technologists around the world, and so what we've built is we've built a technology platform that marries both the physical world as well as the digital world to be able to synthesize materials that can drive those reactions, scale them up and study them in the context of these real world systems called electrolyzers that drive these clean chemical reactions, and we're able to do that experimentation orders of magnitude faster than the status quo and then, of course, like we were just talking about, generate massive amounts of high quality, unique data that we can then apply math to in the form of machine learning and AI, to accelerate this process and really drive understanding, not just of these systems that are being used to split water into oxygen and hydrogen, but now potentially go study things like how can we turn renewable feedstocks into things like the precursors for nylon or the production of hydrogen peroxide, or potentially things like ammonia, the world's most common fertilizer.

Jeff Erhardt: 29:50

So, fundamentally, that's what we're doing.

Silas Mähner: 29:53

So I'm going to try to make sure I repackage this to get into my non-engineering person. That makes sure I can understand this. You're basically you have a method to be able to test out, without maybe necessarily actually doing it physically in a lab. You can probably test it on a computer Many, many different processes and applications of different chemicals et cetera, to identify potential outcomes that might work, that you can then kind of further study and see if it actually is legitimate. Is that correct?

Jeff Erhardt: 30:22

It is correct, with one clarification we are doing this physically in a lab, so we are able to take and not just do simulation within a computer. We are able to do physical experimentation in a lab to study these at a scale that nobody else in the world has seen before. And so the analogy I might give you to think about is if you followed what happened in the world of genomics and drug discovery a decade or two ago, and so the companies that were doing the work with the gene chips, where they would take a chip and they would synthesize a large number of proteins on one of those chips and they'd use it to then run experiments to figure out, okay, what are the best molecules that can drive certain quote reactions in the human body ie, you know, develop drugs is effectively analogous to the technology that we have, except that we're doing it in the context not a protein's, drugs and people. We're doing the context of inorganic materials, chemical reactions and then physical things for the world.

Silas Mähner: 31:30

And so, as part of that process, do you spend a certain amount of time just digitally figuring out, okay, how you know what's close, and then go to spend your resources on actual, you know, materials? And if so, I want to understand how do you guide the computer, the AI, to start, you know, poking around certain areas you mentioned earlier, focusing on certain materials that are closer to?

Jeff Erhardt: 31:56

reality versus you know way out there.

Silas Mähner: 31:58

So I want to understand how you guide it and if I have the correct understanding of digital.

Jeff Erhardt: 32:04

You do. You absolutely do so. Think back to the discussion we were having earlier about the uniqueness of industrial AI and that interplay between the physical world and the digital world, or between physics and data, and the same analogy applies here. The other one I might give you to think about it is if you haven't read it, I highly recommend some of Gary Kasparov's writing the famous chess player from when IBM was developing deep blue and teaching computers how to play chess, and it's a really fascinating story about how these things can play out.

Jeff Erhardt: 32:43

But, in a nutshell, when people first started trying to teach computers how to play chess, they would always lose, and the people the humans would win, and he in particular when he was the best chess player in the world. And then they started getting better and better and better. And they started getting better because there was both more human knowledge embedded within those computer systems, but they also had greater processing power to be able to think faster. And then, all of a sudden, deep blue came and it beat Gary Kasparov, and that was a big, momentous change and everybody in the chess world thought, okay, that's the end, now it's just going to be the computers playing against each other. But what happened was and they figured out you know what, if we could create a team where there's a computer and a great human player, they can do better than either one alone, and so you can think about that as exactly what we're doing. So you can start from the sea. We have great technical experts, people who are deep experts in material science and chemistry and electrolysis, and they spend tons of time studying, reading papers, doing research individually, and that creates the original sea.

Jeff Erhardt: 33:55

So what do we point a first experiment at?

Jeff Erhardt: 33:57

We point a first experiment at their human knowledge and expertise to be able to do this synthesis based on what they think is right.

Jeff Erhardt: 34:07

However, because we can do so much real world experimentation, though quickly, so efficiently and so cheaply, we can also take risks on things that might be wrong, that those human experts wouldn't have predicted can work, and then we can run that all the way through our process.

Jeff Erhardt: 34:29

We can scale those up, we can integrate them into systems, we can understand how they behave in a chemical reaction, and what we do by doing that is generate these massive amount of data sets about what worked and what didn't work, such that we can then create a loop and go all the way back to the beginning and now we can make a better recommendation about what that next experiment should be Not fully based on the computer, not fully based on the human, but on the interplay of those two together. But again, the key point is we're going to do some things that we think make sense and we're going to do some things that we think deliberately might be wrong. And in the sort of machine learning world people think about that as what is my ratio between exploiting something that I think is going to be right and exploring something that might be wrong, because sometimes when you explore something that might be wrong you can stumble into a really interesting discovery.

Silas Mähner: 35:27

So this is. I think it's quite interesting. I'm going to ask what's going to be probably a multifaceted largest question. Let's see if we can get through this, but in the context of what this is as business. So Matique has raised funding and so that means it's not just a philanthropy R&D lab right, it has a business model behind it. So I want to understand that in a second. But I also want you to put it in context of you talked about this idea that you sometimes try to do things that may not make sense, and I want to, if you are able to help non-academics understand how does it typically work in the academia world, where people might be a little bit afraid to try something that could be, that could ruin their career or could not result in something, or it's too out of there, too out of the window of possibilities that they're just not even studying it. So could you take these two things together and help us understand the business model?

Jeff Erhardt: 36:22

Yeah, absolutely. So. We'll start with your second one first, which is the sort of the risk and the status quo of doing any kind of research, which is, by the human nature. Everybody has biases, and one of those biases might be what is the right question to ask? And if, by definition, because of incentives, because you want to get published, because it's only what was published in the past, right only gives a certain amount of ideas, by definition limits the scope of the questions that you may want to explore and in reality it ends up being a very small sliver of what the world of possibilities are.

Jeff Erhardt: 37:09

So the simple answer to your very deep and sophisticated question is you're exactly right, is the risk of letting humans formulate hypotheses based on their own biases as well as the biases of the successes that were published in scientific journals, creates a very narrow set of possibilities? And so you can think of what we're really doing is exploding that set of possibilities out by orders of magnitude and taking the risk to experiment with and explore things that people may never thought would work. But to us the downside of that is effectively zero. But it gives us the opportunity to then understand is there opportunities of untouched potential because people never considered them before because of their biases, and their biases what were published? Does that answer your first question about some of the risks there about the status quo in existing research?

Silas Mähner: 38:12

Yeah, I think it's interesting. It's just something I've thought about and I've heard people talk about, which is this idea that a lot of scientists they're not willing to go for moonshots because the incentives are not there. And even if they did discover something incredible, usually their upside is substantially capped in many cases. That's right, so I appreciate that.

Jeff Erhardt: 38:31

You know you're exactly right, and it's a big and sort of dirty secret about the way that works. So if you wanted to go read some of the stuff about the Alzheimer's research and sort of that perpetuating loop of let's call it not fully robust research that then other people build on top of, build on top of, build on top of, and then all of a sudden you look backwards and you find out the foundation was flawed Right, and so the goal is to not build sand castles on the sand, so to speak, or castles on the sand. So that was that piece and it's very important. You're dead on it. It's a big part of what we're trying to solve.

Jeff Erhardt: 39:09

I think the second question you asked was okay, great, this is all nifty. Right, you can have a big vision about trying to save the world. You've got some nifty technology that allows you to do experimentation, development, exploration, discovery of chemical reactions in the materials that drive them better, faster than anybody else. Great, you're not running a charity. How do you make money?

Jeff Erhardt: 39:34

So the best analogy again that I could think of is go backwards to thinking about the sort of biotech, to big pharma, and you could think about us as being the biotech company.

Jeff Erhardt: 39:45

So, for example, what Genentech was to Roche, right, we could be or are to, the big legacy chemicals producers, whether it's PSF, whether it's 3M, whether it's Dow, people like that. And so our expertise and our ability is the ability to do very rapid experimentation development through what's effectively the pilot production phase, ie scale up development of these systems. But our goal right now is not to become a full-stack or a producer of these end products. I've spent enough time working within those companies, working with them, that I understand. That's an entirely different expertise. There's tremendous amounts of know-how, relationships, channel scaling, reliability for how that's brought to market. So our business model is to develop these technologies again through the pilot phase and then partner with the existing companies in the space to take them into volume production and into market, and then we basically share that success as we take these again more sustainable versions of chemicals and fuels that already exist in the marketplace.

Silas Mähner: 40:58

Sorry, just one clarification there. So you're not just developing the technology to then license the IP out. You're doing this in partnership with somebody else who would then go, take care of it and actually produce the chemicals, correct?

Jeff Erhardt: 41:09

That's exactly right.

Silas Mähner: 41:10

Oh, okay, interesting. So why that versus the IP play?

Jeff Erhardt: 41:14

Yeah, the IP play play is very certainly has some appeal and it's a natural.

Jeff Erhardt: 41:20

it has a natural draw, ie it's very capital, efficient and easy to scale, easy to repeat the reason is is the uniqueness of this space and, unlike IP and some other places where there's a more developed pipeline from the lab through development and into production, this space is not yet as developed in terms of that pipeline and, in particular, there's a pretty big gulf between the lab and production about the way sort of IP meaning materials and processes behave in the lab and how they behave in production. And so there's that gulf creates two things. One is it creates a big time lag and it creates a lot of expense in perfecting these new technologies, and so my firm belief, really ingrained in me from my early days, of taking new technologies, integrating them into semiconductor manufacturing processes, is that you really got to create a tight closed-loop system there, and that only happens working closely together between development and manufacturing.

Somil Aggarwal: 42:33

So there's a very clear parallel there between the operating expertise and the operating excellence and why that factors into the business model. Just to give you some quick numbers here, it looks like you've raised about 15 million in the aggregate. Is that accurate? Yeah, it's close, yeah, close enough. So, with that in mind, one of the things that, when it comes to putting this business model into perspective of, like you said, it's sometimes people would look at having capital off of your books in terms of the investment and R&D is more attractive to scaling. You went through this row and, additionally, we're able to raise on top of it what really factored in. I mean, from an operational point of view, it's very clear why it makes sense from a fundraising point of view. How did you put that into perspective?

Jeff Erhardt: 43:14

Yeah, it's really. Fundraising is a, I guess, kind of like finding a mate. In all of the senses it's still a very inefficient problem, and fit is incredibly important. And so, as we went into this, then the partner that we chose to lead that round has a deep understanding of this space, meaning of physical processes like this, spun out from universities and attacking or tackling very difficult sort of called legacy manufacturing processes.

Jeff Erhardt: 43:52

And so, yes, it was a part of telling the story, telling the big vision, but also coming back to that word we spoke about earlier, which was the realism or the pragmatism and what it truly makes it takes to make these industrial scale processes work. And so, then, it was a matter of both creating that big vision but also being sort of the adults in the room to say, look, we're not naive about what our expertise is and how to make this work over time. And then it was a matter of finding the right partner in this case investment partner who understood that and was willing to be on what's going to be a lengthy journey with us. Thank you.

Silas Mähner: 44:34

Yeah, that's pretty interesting. So I, in particular, when it comes to fundraising, I think it's always fascinating. Everybody brings up the importance of having the right fit, but, yeah, a lot of people are just like scrambling and trying to pitch to everybody, right? I think people need to be careful on that.

Jeff Erhardt: 44:49

So I think it is interesting.

Silas Mähner: 44:52

I just really find it fascinating that you've raised for a model like this, because I mean, maybe I'm just not aware I'm sure there are companies that have something kind of related but kind of as an R&D shop, raising venture funding to go and, you know, actually create something to have some value there. So I think it's interesting. I wanted to initiating their comments. I wanted to shift over to the challenges around commercial industrialization. I think it probably plays into the business model rate thing. So I've heard other people bring up. When you're trying to, you know, come up with a new chemical to put in some machine, you know these big, huge manufacturing facilities are going to be very skeptical to put something in their machine that costs, you know, $35, $40 million, whatever. So can you talk about the challenges of decarbonizing that space and just things that other people working in space might keep an eye out for, things that you've learned?

Jeff Erhardt: 45:44

Yeah, sure, absolutely. You know, I think the first thing in these sort of interplay, with sort of the recap from the last part of the conversation and then lead into this one, you know, whereas you said, gee, you think it would be, you know, challenging to have venture funding for sort of an R&D shop kind of business model, and I agree with that. And what I mean by that is, you know, while R&D is at our core, our business model and our product is to really build products that are then ready to scale into production. And so I think the really what that ties into is that risk factor that you're talking about, which is sort of the skepticism and the needing to see, ie to have it be proven, that these things work at scale In reliable processes, ideally as close to the status quo for the way things are done today, without simply saying, hey, here's a R&D or a laboratory type of answer, you know, go figure out how to implement this yourself. Or coming in and saying, hey, here's what you have to do, you have to, you know, rip out your whole factory to change something, etc. And so finding that sweet spot of how to be able to drive that change management in a way that recognizes there's embedded knowledge, there's embedded know how but there can be upgrades over time and not simply backwards looking rip and replace is really one of the most important things that we've learned. But it is both the refinement of that business model, which you are rightly sort of pushing back on about the flaws in it, which is why we're going on this path, and also recognizing that sort of change management. Again, that sort of patience, persistence in the upgrade path, understanding capital cycles of companies etc. Is incredibly important Because that means for us.

Jeff Erhardt: 47:42

It means for us that one of the most important things we do before we do any work on developing a new process is dive very deep into what's called the techno economic analysis by understanding how feasible is this problem to solve technologically, what economics look like to be scaled into volume production.

Jeff Erhardt: 48:06

And I think one of the things that that I've both seen and heard from a number of different companies that raised money over the last, you know, say, four or five years working in this space, thinking about tackling sort of clean chemical and fuels, is they didn't go that deeply down the economic analysis.

Jeff Erhardt: 48:24

And what's really interesting is is a lot of people I'm hearing this more and more, really raised money during, you know, 2019, 2021, when, you know, money was free and they started doing a lot of development and what they've started to realize is void.

Jeff Erhardt: 48:41

Now I can't get people to offtake this product that I am producing because it's not economically feasible, and that's what this really shows me to do. And again, the reason why we're going through this sort of, you know, partnership route for production, the reason why we're pulling the economics in up front back to before, the pragmatism about starting with the ones that are both more technologically feasible and more economically viable is to effectively solve that problem. And I would say, to put the recap on it is you know, there's a sort of cliche that the best time to start a company is in a downturn, because you're forced to be lean, you don't have cheap capital, and this is the perfect example of that, which is we have to think very carefully about how we make money, what's economically viable, because we can't simply burn through investor capital to deliver on our vision.

Silas Mähner: 49:34

And another thing I want to ask around partnerships, and this is difficult to try to put clearly. I'll do my best, but with developing these new things you've got you kind of get it out of the lab and then has to go through basically a pilot project. I'm assuming you may have I don't know if it's one or is it multiple partners that you kind of work with based off of the chemical.

Jeff Erhardt: 49:55

It probably really depends. So there's certain ones, there's one, certain ones, there's multiple.

Silas Mähner: 50:00

I could imagine that the large companies who are seriously decarbonizing have some budget on R&D already. They probably have some test machines that they're willing to try these things out. So I'm just kind of curious if you have any insights for people looking to try to sell a new chemical or get into these large companies or these large corporates that are already producing things. I've heard some people mentioned that they tried selling to the end consumer, who was actually making the bottles or whatever. Right, but they weren't the one making the chemicals.

Silas Mähner: 50:33

So then they realized they had to go to the different part of the value chain.

Jeff Erhardt: 50:35

Do you have any?

Silas Mähner: 50:36

advice in terms of where to target your sales, or where to target your partnerships for people in the space, or what you guys do as well. I'm just kind of curious on that.

Jeff Erhardt: 50:45

Yeah, I mean, there's absolutely. Here's my advice. My advice is they can all work in different ways and they have different pros and cons, but whatever model you choose, it needs to be self consistent. Okay, so, for example, the people that we already talked about say the pure licensing model into the big companies and that can work for some people. It's not what we're choosing to do. It's got certain pros and cons that we talked about.

Jeff Erhardt: 51:19

On the flip side, the arguments of going straight to the end consumer is really related to the ability to get that brand recognition and start to be able to sell something as quote more sustainable, right, and the bet that those companies are making is that their brand recognition and that the end consumers will pay that premium because the process is more expensive and that's okay.

Jeff Erhardt: 51:45

That might work and there's certainly debates about that and they change depending on economic conditions etc. You know, as a sidebar there's there was an interesting article about all birds I think it was in the Wall Street Journal, right about that. You know them trying to go down this path really sells sustainability etc. And even for something as consumer facing as shoes. That was really challenging and they faced really business challenges because consumers just weren't willing to pay for it. But my point is that's okay, you can try that model, but that means that your marketing, your branding, the way position yourself, that focus needs to be tied to that. And so what we've chosen to do is to solve the problem of sort of the green premium, not by asking people to pay more, but by making our processes more economically viable and efficient on their own, okay, such that people shouldn't have to be willing to pay more simply because it's sustainable, but they should be willing to pay more because it's better for other reasons.

Silas Mähner: 52:54

Yeah, that makes a lot of sense.

Somil Aggarwal: 52:56

And one thing you know, I, I. This is always an unfortunate part of the show, but we're coming up towards the end. I think you're especially the way where you operate. You're in a unique position where your business model puts you in contact with a lot of emerging areas within your space, yeah, and the ability to see sort of on the ground, almost almost like like a partnership model, but in this case that you know it's it's more beneficial in terms of revenue and the kind of things that are coming about. So, at a high level especially for the aspiring entrepreneurs who will take away from this conversation that there's a real industry here. It's nuanced, takes a deep understanding, but there's a real industry At a high level. What are the opportunities that you're seeing to build companies just based on, especially, your wide reach within the industry?

Jeff Erhardt: 53:40

Yeah, I mean, look, there is absolutely a gigantic opportunity here. So if you look at the broad space, you know sort of chemicals, fuels, productions again, it's hard to abate sectors which would also include, you know cement, you know secret steel. Obviously we're not addressing those. Those are incredibly important, they're incredibly hard and they are giant market. And so even for this one, you know this idea of how do we do clean chemistry Again, whether that chemistry is the electrochemistry, whether it's the synthetic biology which is there's people thinking about solving similar problems in that way.

Jeff Erhardt: 54:16

Really, what I would encourage them is to you know, bring your expertise that there is tons of room for innovation, for new companies to be doing this, because it is a giant market. There's tons of opportunity and eventually, you know, we'll start to see things shake out at predict in the next, say, three to five years about which technologies are best suited to solving which portions of this. Let's call it clean and sustainable fuel markets. And so all that I would say you know to them is yes, bring your technological capability, because there is tremendous opportunity. However, I'll repeat what I say to our own team all the time, which is just remember that a technology is not a product, and a product is not a company, and so what I mean by that is, as you think about developing your team, figuring out what markets you're going to go after, think just not about the raw technological expertise, but how you package that into a product that makes sense.

Jeff Erhardt: 55:24

The questions that you guys were asking okay, do you package it to go full stack to an end consumer or do you package it as R&D play? Those are okay, but it needs expertise and debate to be mutually self or internally self-consistent. And then, finally, the last piece is from the company standpoint, which is, what is your sort of expertise? How do you figure out what you guys are hinting at, which is, what's your go-to-market? How do you start to deconstruct the value chain? How do you figure out, are you going to go compete against the big guys or are you going to play with them? And all of those need to come together. So I would encourage people if they're thinking about this. The world needs more of it.

Silas Mähner: 56:03

I think it's interesting. We had Matt McGrath from Anthro recently and he was talking about how 75% of the world's largest companies are all hardware companies. I think it's important to people to remember that especially. Yes, people usually paint them as the bad guys, but they are going to be the ones, when they decide to decarbonize, that are going to have the money to invest in these technologies and decarbonize all of these existing assets 100% that stuff's not going to go away and that's really the critical point.

Jeff Erhardt: 56:34

You need to go back full circle. You asked how I got into this then. Why Look? I live my main home in San Francisco. I've lived in the Bay Area for 25 years. I hang out with software guys. I've done two software startups. I sold one to Microsoft, I sold one to GE. It's interesting. We always think the Mark Andreessen quote of software is going to eat the world. Yes, to get outside of that bubble, it's the physical things that still build power and, unfortunately, create emissions in our world. The ability to tackle those and play in that sandbox and be pragmatic about it, I think is incredibly important, not just for us as a company but for anybody else who wants to get in this space.

Silas Mähner: 57:19

I think we've got time for one more question. It's a little bit self-serving because, since I work in the town, space in my day job, but I'm curious, especially when it comes to hiring very technical, especially PhDs on different, especially in this case, chemical engineering. Most of the time. Do you have any advice to people, to companies who are hiring, Because I've seen a lot of new climate founders who are very green They'll say we have this really critical hire but we're not willing to spend any money on helping find that person. We're just going to go through our networks. I've seen some people who have been through startups and who have realized, hey, this is probably really important, I should hire somebody to help me find the best person in the world for this, because maybe they're going to bring along their followers because people are usually looking up to some of the other scientists. Do you have any advice when it comes to hiring PhDs, in particular very, very technical people?

Jeff Erhardt: 58:12

Yeah, a couple of things I would say. Again, the first thing I would say is know what you're good at. If you're a technical founding team that knows how to evaluate and has the network to be able to do that for your technical hires and again, this may be anti-serving to your self-serving question I wouldn't go that route. I would say you have to do that. But if you're a non-technical team and you don't have the network, I would say absolutely leverage somebody who has that network and expertise to be able to pull some very good technical people in. I'd say, conversely, on the business side, if you're a technical team, I think there is tremendous opportunity in leveraging high-level recruiters for key leadership positions. Here's what I'll bring that now back to the technical side, which is so.

Jeff Erhardt: 59:04

The model I've used in the past is a couple of things.

Jeff Erhardt: 59:05

So, first is, know what you're good at, whether you're using outside help or whether you have the expertise to do it internally.

Jeff Erhardt: 59:13

The way that I've thought about doing this, or the way that I still like to do it, is to start with somebody who's getting an icon, somebody who young talent will want to work from, learn from and be excited by Number one and then from there start to build the machine or the talent pyramid, not quite, as say, aggressive as what consultants or investment banks or law firms do, but start getting the ability to develop and attract your own talent based upon a couple of these iconic people that the younger people want to come work with and learn from. If you can get that machine humming, both how you interview, how you develop them internally, how you go earlier into the pipeline and start to have a structured intern program which is effectively an extended interview process, you can start to scale things very well and very cost effectively. And so, to bring it all back together, you got to get that first one right in a key role that then access that nucleation point that you can build off.

Silas Mähner: 1:00:25

You said better than I could. This is exactly the point. I think most people are too afraid of the upfront cost, but they don't realize the subsequent savings when you can do this properly. So I appreciate that it's a very fun topic. I love the talent space. It's hard to find the right people.

Jeff Erhardt: 1:00:42

That was an understatement.

Silas Mähner: 1:00:44

So, anyways, this has been really great. I really appreciate you coming on. Any final thoughts for the audience things you're looking forward to Call us to action anything?

Jeff Erhardt: 1:00:51

No, absolutely. I think it's a really exciting time. There's a lot of stuff going on. We have a lot of very interesting projects going on. Stay tuned for some announcements over the next couple of months. We'll have some things coming out with some pretty big progress and, I think, some things that will shake up the world pretty good in the production of sort of the hydrogen space, and some of the work we're doing there is going to start to create some ripples in the very near future. So stay tuned. Lots to come and look, we're hiring too, so feel free to reach out anytime.

Silas Mähner: 1:01:23

Awesome. Thanks so much, Jeff. We really appreciate you coming on.

Jeff Erhardt: 1:01:25

You're welcome. See you guys.