 Good afternoon, can you hear me? Good afternoon and welcome to today's Energy Salmon. It's really great to see you all here in person, pleased to meet you, kind of. Today, our speaker is Vivas Kumar. I'd like to introduce him this way. We all figured out that the last time this class took place in this room was March 2nd, 2020, and our speaker was Vivas Kumar, then a, I think you were affiliated with Stanford Business School and a kind of strategic partner at Benchmark Mineral Intelligence, which is probably the leading firm in supply chains for materials, particularly in that talk, Lithium-Ion batteries and whatnot. He gave just a dynamite talk, it's actually online, and he says he gets a lot of YouTube hits on it. And then he recommended the managing director of Benchmark Give-A-Talk, and that happened last year from London on April 26, 2021. So I didn't think we would hear from him, and then we were talking to Will Chiu, some of you may know him, professor here, extraordinaire in material science, and he was very one of the leading battery materials guys in the world, and he was the co-director of the research program called Stuart Jax, and he said, guess what? Guess what Vivas is up to with us? A brand new startup, not long ago, called Winthrop Chem. So he's gone from being the world's expert on supply chains, he was prior to the Benchmark job at Tesla, which has done a pretty good job also in that area, and he's kind of transitioned to another big piece of what's going to take to do the energy transition we've got, and that is getting new technology from labs to market production, I guess is your term for it. And now he's going to tell us how we can reduce that time lag, which is usually pretty onerous, and has risen a lot of promising startups out of business, for example, in something like 80 or 90% less time. So Vivas, tell us how you're going to do that. Thank you, Sarah. Thank you, everybody. It is really good to be back. 690 days ago, I gave the last in-person Stanford Energy seminar standing in this room at this exact same spot, and my life changed forever because of giving that Stanford Energy seminar. I didn't cause the pandemic, but there is a direct line you can draw from the fact that I gave that Stanford Energy seminar to having started this company that I'm presenting to you today. And by the end of this presentation, you'll be able to see exactly what I mean, and why it's not an exaggeration to say that standing in this spot at 4 p.m. Pacific 690 days ago changed my life. It's really going to be back at Stanford also because, you know, the ethos of Stanford, I was a student, I was in your shoes just eight months ago, is to provide an educational experience that lasts a lifetime. And when I think about Mitra Chem, for me it's the culmination not only of the couple of years that I was here at Stanford, but the multiple years before that that I was doing battery work at Tesla, the years that I spent at Stanford also doing work at Branchmark Minerals and thinking critically about how batteries are the core platform technology to enable electrification is one of the most important technologies to help us fight climate change. And there are many unsolved problems that need entrepreneurial energy from the types of people who are sitting in this room. So what I hope is that after today's talk, all of you who are in this room will be inspired to think about how you can also have a role to play in solving climate change, in perpetuating a lot of these technologies that need to be deployed and see the light of day and see wide-scale deployment across boundaries, across geographies. Now, for those of you who are sitting at home and watching me on YouTube right now, last time I spoke over here, there were 100 people in the room and I think I got several thousand views on YouTube. I know that many of you are here just to hear about Mitra Chem. Well, I am here on the invitation of the students to be here. For the first 10 minutes, I'm going to give a pretty generalized lecture about platform technologies and supply chains. If you are here just to hear about Mitra Chem, skip 10 minutes, look for this slide where the second point on this slide is highlighted in green. That's what I'm going to talk about Mitra Chem for the rest of the time. But for the rest of you who are in this room, let's first talk about what it means to be a technology platform. So, we think about technologies in terms of the stack. This is a concept that goes beyond multiple engineering disciplines. Assembly language is made up of, or rather coding languages are made up of assembly language is made up of switches. Semiconductors are the input material for computing chips, which is the input material for computing. And likewise, we can think about electrification and batteries as falling within this technology product stack as well. Namely, that when you have categories, these are largely immutable categories. They are related to how human beings essentially function and grow the consumption of energy, the consumption of water, the consumption of healthcare going into the proliferation of biological sciences. Categories are unlikely to change over time, but what are likely to change and develop over time are the platforms and technologies that are sitting on top of these categories. So, let's just take a clean example over here. When we're thinking about energy storage, earlier I mentioned the lithium ion battery is a key platform for energy storage. And there are multiple product categories built on top of lithium ion batteries. For example, consumer electronics, grid-scale energy storage, power tools, electric vehicles being the most notable public example. Now, what do you do with an EV? There's many things you can do with an EV. It's a 2,000-pound machine that you can use to block somebody's way. You could use it to flatten something. But in reality, for the most part, you're using it to drive. That is the primary function of electric vehicle. And that is the main product that you're using it for. Now, a feature that aids the lithium ion battery is this idea of regenerative braking. So, how can your deceleration lead to the recharging of the battery at the same time? That is one small part of the way in which a battery works to discharge electrons and to recharge itself within the overall product stack of an electric vehicle. Sometimes this slide can be presented with products and applications switched. There's a big sort of war in terms of product thinkers and system thinkers about whether products supersede applications. But the reality of the matter is we are thinking today about platforms and how platform technologies enable us to solve problems like climate change and proliferate technologies through electrification. Progress on these platform technologies is measured based on picking certain variables and either maximizing or minimizing them through recursive functions. Positive feedback loops or negative feedback loops. Here you see examples of three well-known platform technologies outside of batteries. So for semiconductors, we are sitting here in a building. We're sitting here in the auditorium, the NVIDIA auditorium. NVIDIA is one of the largest semiconductor companies in the world. And what are they doing every single day? They are trying to make sure that the number of transistors on a GPU chip increases while the cost of that chip and the power consumption of that chip is continuously decreasing. Likewise for biotech, what you're optimizing for is saving as many lives as possible while minimizing the potential of side effects. So maximizing a variable that you like, minimizing a variable that you don't like through iterative compounding gains, experimental cycles that go over and over and over again within individual companies, within individual product categories, within individual applications. Three important laws have been used to generally govern the way you would think about platforms. And these three laws are most associated with semiconductors, although they can be associated with other forms of platform technologies. For example, if you take these three semiconductor laws and think about how they apply to batteries and why they may not apply cleanly to batteries, it is true that while a industry is experiencing hypergrowth, your cost will generally decline as a function of time, or in the case of Wright's law, your cost declines the more and more units you produce and the more and more of a learning curve there is in manufacturing. But what we are seeing right now in the battery industry is quite interesting. We are starting to reach diminishing marginal returns for some of these power laws of platform technologies. For example, it is no longer true that cost will just exponentially decline for lithium-ion batteries in the way that they have over the last 15 years. Your average lithium-ion battery, average, global average, has gone from approximately $1,300 per kilowatt hour, excuse me, all the way down to nearly $100 per kilowatt hour over the last 15 years. Tremendous innovation and growth. That is what has enabled so many electric vehicles to be on the road, but we are not going to go from 100 to 10 anymore. There are certain physics-based limitations, especially as it comes to the pricing of the materials used in that battery that come up against the reality for this power law to continue over time. Likewise, the explosion in the number of plants, we went from having two battery factories in 2014, all of doing more than one gigawatt hour per year of production, to now having several hundred battery plants, either in construction or already producing that can do more than one gigawatt hour per year of production. But that doesn't necessarily mean that the supply chain underlying those plants have also been built. And so what is the potential of the cost increase on an individual unit basis, but also for the plant? That too remains to be seen. Finally, on this idea of supply chains, what we have seen over the last two years due to COVID is an increasing emphasis on the study and the idea of managing your supply chain as being a core value driver for businesses. These are the four themes that have emerged as being most important for corporate executives, for boardrooms, and also for supply chain managers. Having a supply chain that is resilient, that is multifaceted and is able to withstand the shocks of interruptions at various places in the world, a supply chain that is compliant to the environmental, social, and governance standards that are expected of them, and coupled with the transparency for tracking and traceability, and finally a circular supply chain that allows a product to amortize its carbon cost of manufacturing over a longer period of time. Battery supply chains tend to be very complicated. This is as simple as you could present it, but in each of these steps are multiple companies that produce the materials that produce the reagents that produce the equipment needed for manufacturing. And here are examples of today's producers, but there are several other companies that are entering the space and several more relationships interlocking the supply chain that hasn't yet been explored. If you want to learn more about this topic, go watch my Stanford Energy Seminar for March 2020 on YouTube, where I explain all of these steps of the supply chain. But today we are going to focus mostly on the part that's highlighted, the electrode supply chain. The last point that I will make about platform technologies is what happens when you don't plan for the growth of platforms. Something like this. One of the most notable news stories over the last two years has been the semiconductor chip shortage. This is still an ongoing problem that involves the world's largest governments and largest companies. This is a warning shot to the battery industry. If we don't build the supply chain for batteries and battery materials and the machines needed to build these machines, we are going to see these same headlines about the battery industry within five years or less. We need existing producers and new companies. New companies like the one that I am now part of building, which is MitraKem. The goal of MitraKem can be summarized basically in this one slide. We aim to be the first lithium-ion battery materials product company that shortens the lab to production timeline by over 90%. Today we are focused on phases number one and two, but over time we will be going through and doing all of these phases, inventing new category of high energy cathodes and manufacturing at scale, taking advantage of an in-house R&D acceleration capability. The one word that is underlined on this slide is product. The most important word. We are going to be a product company from day one. It is not sufficient to solely accelerate the process of innovation, but rather using that as an in-house advantage to build products that customers want and then industrializing those at scale, putting cathodes into batteries, into cars, onto roads. Which begs the question, what is our first product? Our product thesis is this. What a customer wants is to have a car that meets the specifications that they desire, not the specifications dictated to them by the technology that is provided to them. When we went out and surveyed automakers and battery makers about what would be in the Western world the mass market battery technology of choice, from a first principle standpoint, thinking about the metrics that matter and not reasoning based on today's existing chemistries, we came up with this sweet spot. A battery chemistry that can sit somewhere above 200 watt hours per kilogram for gravimetric energy density and at a cost for cathodes that is comparable to today's leading LFP solution. Now what do these letters mean? In today's market there's a bifurcation between high nickel rich cathodes and iron rich cathodes. Over the last 10 years on a simplified basis the game in the West has been take your nickel rich cathodes and push it up to get to number two on that page. The game for LFP has been to push it down on price as much as possible through commoditization. And today automakers and battery cell makers are forced to pick one of these two existing chemistries. When in reality nobody has yet industrialized a solution for a cathode that can sit in this sweet spot. So there's three ways to get to the sweet spot. Number one, you invent a completely new cathode material. Why doesn't that work? Because automotive qualification processes are very, very long. To have your material be fit for purpose and fit within the architecture of vehicle programs for a completely new and unproven material will take over five years. That is antithetical to our stance that we need to get product into cars and on the roads as quickly as possible. So the other way to do it is to take the existing solutions and try to move them into the sweet spot. Why does it not make sense to take a nickel rich cathode and pull it into the sweet spot? You run into the reality of the volatility of nickel and cobalt pricing from the London Metal Exchange. There's really no way to completely control for all of the material cost variability that's happening. And so the other way to think about it is you take the category of iron rich cathodes and try to push it up into the sweet spot. Ultimately, keep pushing, keep pushing through the compounding incremental gains that we have seen in other great platform technologies and push it all the way up to category one. This is the product thesis for Mitra Chem. To solve the sweet spot and ultimately over time get the category number one. And our first customer for whom we are doing this, the first customer segment is mass market electric vehicles. Mass market electric vehicles are the largest and fastest growing market in the world. When we looked at other great platform companies, one thing that we saw, especially with Taiwan semiconductor manufacturing, which as many of you know is one of the largest companies in the world and is the biggest semiconductor fabricator in the world, their first product when computing was seeing the same revolution that we are seeing today with electric vehicles was to pick the largest and fastest growing market and sell product directly into that market because you are not trying to displace existing suppliers as much as you are growing with the demand of the industry. Going back to automotive qualification, displacing an existing supplier is frankly an exercise in futility when you are a relatively small company like us. But rather growing into supply and growing alongside the tailwind of the industry is a much sure bet to get to cash flow as quickly as possible. And we are seeing that right now, the undeniable growth of electric vehicles. Ever since I spoke at the Stanford Energy Seminar two years ago, this trend has only accelerated. More and more companies are coming out with new electric models. I'll give you another example of this. Ford, an iconic American company, initially said that they were going to produce X number of vehicles for Ford F-150 Lightnings. And then two months later, it was 2X. And then two months after that, it was 2X that because they just couldn't keep up with the demand. So the demand tailwind is a great tailwind for our business as well to just follow along the growth of the broader mass market electric vehicle industry and to take advantage of the demand pull and the technology push. Going back to the point I made earlier, when we scale our industrialization, there will be low initial yield. But it can be offset by selling it to the largest and fastest growing market as possible. And what that allows us to do is generate cash flow to create recursive, positive feedback loops within our own R&D programs within the company. So we are betting on five main themes. And these are the five themes that dictated why this product strategy and this market strategy is what we are going after. The first is batteries are here. They are the platform technology of choice to power electric vehicles. Yes, there are certain vehicle segments for which other energy storage mechanisms like hydrogen can work. But for the most part, for the majority of the mass market applications of electric vehicles, we are seeing that lithium-ion batteries are here and we are seeing a commensurate growth in cell capacity. And to be frank, this chart will probably still keep growing exponentially. The only reason it cuts off and levels off at 2030 is because we just don't have enough data to forecast any further than that. And the truth is, as lithium-ion battery capacity grows, so too does cathode capacity, because lithium-ion batteries by definition have a cathode and an anode. Those are the electrodes that make up the lithium-ion battery. Cathodes are the most expensive part of manufacturing the lithium-ion battery. And there are four characteristics that matter the most when deciding the efficacy of a battery. Cost measured as $1 per kilowatt hour. Energy density measured as what hours per kilogram. Cycle life, so the number of cycles until your battery reaches end of life. And finally, safety. There are many ways to assess safety, but the easiest and most crude way to think about it is what is the probability of a thermal runaway event happening, where your battery catches on fire and causes a chain reaction for the whole pack. The cathode is the biggest influence across the dance of optimization across these variables. If you change one variable, you're going to get an offset on something else, and vice versa. The cathode is part of an overall system with the anode, with the separator, with the electrolytes, with the physical dimensional characteristics of the battery cell as well. By no means does the cathode spell the entire battery cell's performance, and the interaction of cathode with those battery materials matters as well. But the cathode performance matters the most in dictating the battery cell's performance. So what does that mean? Well, what that means is choosing the appropriate cathode is extremely important for an automaker. We are betting on iron-based, not only because we believe it's the fastest way to get to that sweet spot, which is the optimization of gravimetric energy density and cost, but also because, frankly, we are concerned when we see what's happening in today's western industry and their over-alliance on nickel-rich chemistries, especially because nickel chemicals and cobalt chemical shortages are imminent. I have been saying this for years now, and you can look back at my last energy seminar, where I presented basically this exact same slide, and frankly, the situation has not changed. What we need is more capacity for these specialty chemicals being built. What we are seeing is the exact opposite. More battery cell plants being built, more vehicle facilities being built, but not a commensurate growth in the materials and the specialty chemicals needed for those cathodes. Well, but you can make the argument, we've got secondary supply on the cobalt chart. Surely recycling can make up for this shortage, which I would say recycling can make up for some of the shortage, but recycling as an industry faces its own pressures as well. If we think of a simplistic black box diagram for how the recycling industry works, the input feedstock for a recycling process is used battery cells and scrapped from battery factories. What you need to make your recycling process economics to work because the customer will only be willing to pay the market price for those materials is to ensure that your recycling process can reach sufficient scale such that your operating cost of manufacturing recycled materials is at or below that of virgin materials. To do that, you need to have the largest input feedstock possible. And the issue is that the desire from recyclers to have the largest input feedstock possible comes up against the realities of the battery industry. While the number of battery plants is scaling quickly, as they reach scale, so too does their utilization rate grow, and the battery scrap produced by these factories does not grow at nearly the same rate as the battery factories themselves. Likewise, for used cells, the question now is how do you amortize the value of the carbon cost to make that battery cell over as long a lifetime as possible, which is why secondary cell use, taking all electric vehicle batteries and putting them into energy storage devices, is in vogue and multiple geographies. So recycling is a great idea. There are many companies out there working on battery recycling, and even if everything I said today didn't pan out and recycling economics made perfect sense and beat virgin materials operating cost of production, you still cannot make up for those shortages solely based on recycling. So the other option then is just to shift away from the use of nickel and cobalt. Lithium is going to be in shortage. It is a fact that lithium chemicals are already in shortage. As of this week, we are seeing the highest ever prices for lithium chemicals in the world, and this trend of lithium prices being high does not seem to abate anytime soon. It is a lithium ion battery because it uses lithium chemicals, and the challenges of lithium chemicals are going to be standard no matter what chemistry you choose. But in my mind, why then add on nickel and cobalt challenges as well? From the standpoint of PAC economics, we also would argue that an iron-based cathode tends to make more sense, and here's the reason why. So let's take a simplistic model over here of using nickel-manganese cobalt at 50-30-20% ratios, respectively, and use that as our normalized baseline. The argument will be made by those whose views iron-based cathodes skeptically that your gravimetric energy density makes us touch that the number of LFP-based cells you have to use is about 20% more in weight than NMC, and they are correct on a cell level. But when we look at it from the standpoint of PAC economics, we see PAC sees almost a 33% reduction in the weight. Why is that? Well, the reason is because a nickel-rich battery cell chemistry is more likely to experience a thermal runaway event. And as a result, the PAC has multiple safety mechanisms built into it that add weight. So mechanical features, interstitial fills, flame retardant materials, all of these add to the PAC weight. You get rid of all of those safety features. You can actually pack more cells in, and you layer on top of that some of the hypotheses that we are pursuing in order to push the energy density of iron-based cathodes up, even if you get only a 10% improvement on gravimetric energy density. As compared to today's state of the art, you are still falling 4% below the normalized kilogram per kilowatt hour. If we look at it from a cost standpoint, it is a fact that LFP cathodes today are cheaper than NMC cathodes on a per kilowatt hour and a per kilogram basis. And even if we only charge the market rate, even if we charge no premiums for the fact that our product is an enhanced product, we would still be at a 20% discount in terms of cost to today's nickel-rich cathodes. So if you're an automotive executive or you're a battery cell manufacturer executive, it's pretty clear that LFP on a PAC level can be very competitive for Western mass market solutions, even for the discerning customers in the West. And we're seeing that happening right now, a tremendous shift towards incorporating more iron-based cathodes within the product roadmaps of companies that previously thought that the extreme commoditization of iron-rich cathodes made it a secondary option or a backup option at best, to now being the primary option of choice. And we are very happy to have chosen iron-rich cathodes for this reason, but also, most importantly, that in addition to the weight efficiency gain and the cost efficiency gain, iron-rich cathodes are also safer. Thermal runaway is completely defined by cathode choice. Due to the crystal structure of LFP, LFP cathodes are safer, have a lower probability of thermal runaway, have a lower probability of a catastrophic event happening, like a runaway fire that bursts out of control as compared to a nickel-rich cathode. Switching gears a little bit to thinking about what the automakers want. What the automakers want is not a one-size-fits-all cathode. I mean, I can sit here and put up the slide and talk all day long about how a mass market EV benefits from having an iron-rich cathode. But the truth is, the automakers, they're not just making a mass market car. They're making multiple different types of cars. And what they don't want is to be told that an off-the-shelf cathode is the right solution for them. So this slide actually comes from an experience that I had back when I worked at Tesla. In one of my first ever meetings with Elon, he sat down and told us, today we make only premium luxury sedans, but five years from now, mind you, this was 2016, so he said five years from now we're going to be making mass market cars, premium cars, commercial vehicles, energy storage products, sports cars, pickup trucks. We should not be using the same battery cell chemistry for all of these applications. We need to figure out what does a customer want in terms of characteristics for how their vehicle will perform, translate those characteristics into a specification, design, build, test, and industrialize that specification so that the way in which the vehicles perform are differentiated by application. Today all of these EV manufacturers are still taking off-the-shelf chemistries in order to accomplish these tasks, but increasingly what we want to see is a future paradigm in which the automaker, the cell maker, can define specifications that they challenge the industry to meet. The limiting factor behind being able to do that right now is the fact that it takes so long to invent and perfect a new cathode chemistry. If you can solve that challenge of speeding up the design, build, test, and industrialize cycles, you can get to a faster solution over here, and that is exactly what we are doing. Our in-house advantage, our technology advantage, is the increased R&D speed that turbocharges our overall product innovation. It is not simply enough to have a strong product thesis around iron-based cathodes. For us we also paired that with an R&D acceleration thesis that goes from designing to building to testing to industrializing. Now the way that we've done that is within our lab we can make cathode, test cathode, put the cathodes into cells, both coin cells and pouch cells, test the cells using the exact same testing methodology that an auto OEM or a battery cell manufacturer will dictate at the materially relevant quantities that they would like, run metrology, and then start the entire iterative design build test process over again, all under one roof, all within 10 steps of each other. And layered on top of this design build test cycle is a physics-based machine learning model, developed based on the work of my co-founder, Professor Will Chu, and my co-founder, Chiru Gopal, who were both here at Stanford to do that work. And what that physics-based machine learning model does is it adjusts data about the testing methodology and allows us to accelerate the in-house qualification process to assess whether a material that we are trying to bring to market reaches the specifications desired by a customer in much faster time than is done by today's industry standards. Today's industry standard for design build test cycle is anywhere from 12 to 18 months. We are going to bring that down to one month or less by the end of this calendar year. And what that does is it allows us to go out and test multiple product hypotheses and deliver multiple products to the relevant segments that need them. Finally, coupling that lab-based and pre-pilot-based design build test cycles is having pilot capacity close by us. And this is where we are planning to get to also with our next round of milestones that we're pursuing. It's being able to go from one shot lab to pilot all within the same workflow, all within the same geography, all within one month. Layering on top of that, what do you do when you get to pilot? What do you do when your customer wants to qualify your material? Well, you have to produce it at a commercial scale to be commercially relevant for them. And this is where we had to ask ourselves, well, how are we going to pursue the commercial scale production of these materials? Now, I'll be totally upfront and say we're not ready to build a $700 million factory yet. But the day will come, and when that day comes, we want to do it right here in North America. The geopolitical reality of complex supply chains right now is quite scary. Trying to centralize and create vertically integrated supply chain hubs is one of the best ways you can defend against the resiliency issues that we've seen. And the geopolitical tensions between the US and the China have been well documented, and unfortunately it doesn't look like those tensions are going to abate anytime soon. In light of those geopolitical tensions, it's quite a scary situation when you see that all of today's iron-based cathode manufacturing happens in China, and there is almost zero cathode capacity of any kind here in North America. Save for one pilot plant that's running that does not produce commercially relevant quantities. This is a massive Achilles heel for American industry right here. It has been well documented and well discussed on Capitol Hill, on Wall Street, and here in Silicon Valley. But we have to go solve this problem and coupling innovation with a product thesis with thinking about scale-up from day one is what is needed to go out and solve this problem. This is a problem that goes beyond cathodes and also extends to other materials needed for the energy transition. When we look at copper, nickel, cobalt, rare earths, and lithium, it's clear. Companies operating in China have the significant proportion of the production of most of these materials. We will not be in a position to solve the cathode capacity problem and solve the supply chain for our cathodes. But we are thinking critically about how do we engage with players within the West, within North America to create those vertically integrated supply chain hubs such that this picture also starts to shift and starts to be more balanced to reflect the reality of who consumes electric vehicles in the European Union in North America versus in China. Executing on these theses allows us to see the virtuous cycle of product innovation and cash flow that has defined every single great platform technology product company in the last 100 years. We will start by scaling production for mass-market electric vehicles for iron-based cathodes. Then we will move to qualify with a customer. Although we can do quick qualification loops within our lab, our customers will still demand that we follow their qualification process to get our material in their car. It is a commercial reality and we have accepted it. So we will work with them to qualify our material to get into the supply chain of electric vehicles that are going to be in the Western world. And then after that, once we qualify, we'll begin production scale-up to generate base load cash flow. The metric of success of this business is reaching cash flow. It is not valuation. It is not revenue. All of the other metrics that we track is towards reaching cash flow. I was across the street at Stanford Business School for a couple of years and the biggest lesson that was hammered into me by my classes over there is the metric of success of a long-term successful company is sustainable and resilient cash flow sources. We're thinking about that from day one. Finally, once you've generated that base load cash flow, we can then reinvest that cash flow into R&D for more application differentiation to touch more product segments with iron-based cathodes and potentially with other future chemistries. Now at this point, things get interesting because we can grow new production lines with the new application-specific product and just keep repeating cycle 3 through 5. But what gets more interesting is not only to work on speeding up the design-built test cycle, but also speeding up the industrialization cycle and the industrialization innovation cycle. Putting that into more plain words, it means improving the yield on large factories with those compounding operational gains that you see from process optimization of large-scale facilities that are pumping out tens of thousands, hundreds of thousands of tons of material. And finally, use that improved yield and reinvest that cash flow into future chemistries and products. This is the same virtuous cycle. Apple with Consumer Electronics, Genentech and Biotech, Arcelor Metal in Global Materials Production, Amazon with Market Places. They all follow this virtuous cycle of cash flow and product innovation. And we are now on day one, but we are already thinking about how do we generate this cycle for ourselves. Finally, since this is a room full of Stanford students, I thought I'd talk a little bit about why speaking at Stanford Energy Seminar on March 2nd of 2020 led me directly to being over here. That day was the last day that we could all gather in classrooms on campus. I spoke at the time, it was a fully packed audience. It was well over 150 people. It was wonderful. I spoke from 4 to 5.30 p.m. at 6 p.m. Mark Tessier-Levin sends out an email saying, we're canceling everything. Like, go home. Everything is canceled. Everybody go to Zoom. And I was very lucky that my lecture was in person that day because through that lecture and through being in this room, I got to meet my now co-founder, Professor Will Chu, who's a professor of material science. I see a few of you shaking your head so I'm sure you take classes with him. So he was sitting in this room. I was talking about some of the challenges that I see in the battery materials industry. If you were to go back to the lecture I gave in March, there is a slide where I talk about how automakers crave application differentiation. And he said, hey, I think I know how to solve that. So the next day, we met at the faculty club and talked about some ideas on application differentiation. He said, look, this all sounds great. The lab work that I do might be fitting into the commercial thesis that you have. Let's keep in touch about this. And then COVID really shut things down the week after that. But, you know, Will and I reconnected when I got back to Stanford six months later and he looped in Chiru, who's now our third co-founder, who used to be a former postdoc of his who had also worked on this ML infrastructure work. And after several months of thinking about these ideas, we decided that we were going to start this company. It's also particularly meaningful that today, January 24th, 2022, is exactly one year and one day since we officially decided to start this company. I remember the conversation January 23rd amongst the three of us and with our lead investor for our first round saying, you should start this company, you should write a business plan, you should go do it and I will fund it. And we were off the races at that point. So coming into Stanford, I've heard a lot about how there's such a wonderful entrepreneurial ecosystem here. I am living proof, we are living proof that Stanford's entrepreneurial ecosystem can support you in your endeavor as you think about how you want to play a role in solving climate change, especially if that role involves taking an entrepreneurial path. We went from being three co-founders working on Zoom late at night to now building a lab only 10 minutes away from here in Mountain View. We were very lucky to get this building that has a second floor office space and a first floor lab, which means that we can continue to work in there while construction is happening. We started large scale demolition activities today, so it's quite loud in the office, which is why I'm happy to be here and not there. But the excitement is very palpable. One thing that I really missed after the last Energy Seminar was the excitement that I had being with my fellow Stanford students. I never made it back to the classroom before I graduated, but now being in the office with our team, our wonderful team members who join us from all walks of life and all different companies and all different backgrounds in material science, chemical engineering, machine learning, software engineering, and in company leadership. It evokes the memories that I had of what Stanford was like before COVID hit and the positive reinforcement that I received from the students around me. My sincere hope for those of you sitting in this classroom is that you do take advantage of the peer relationships that you have, take advantage of the professor relationships that you have, be nice to your professors. They may be your co-founder one day. With that said, I would love for any of you sitting in this audience, or any of you watching at home, to come join our team. At the end, I'll put the slide back up, just point and shoot at this QR code. It'll take you directly to our careers website and you can see exactly what open positions we have. Having worked in climate is not a prerequisite. Having worked in batteries is not a prerequisite. All we care about is that you are motivated to solve the defining challenge of our time. Climate change requires a multi-faceted, multi-disciplinary approach and there is a place for anybody motivated by that challenge within our company. Finally, hold me accountable. Two years from now, I'm going to come back to Stanford Energy Seminar. I'm going to put this slide back up and I'm going to tell you what we did to make LFP go up to the sweet spot and how we're thinking about going to quadrant number one, which is the holy grail of cathodes. High energy density at the lowest cost possible. Hold me accountable. With that, I'd like to say thank you. I will open up the floor for any questions and it's been a real pleasure to be back at Stanford today. So anybody want to shout out a question or two? Sir? It seems like this machine learning approach is kind of at the heart of your R&D. I'm wondering, can you talk a little bit more about how that's been going? Was it hard to implement machine learning into R&D and how much has it helped? Sure. So the way in which chemicals companies within industry today generally approach applying ML to R&D is they will outsource the ML function to a generalized team of engineers who may not understand the underlying physics hypotheses and who are not there to see the experimentation happen, which is why I'll reemphasize that the important step that we took was to co-locate all of these steps within the same building. Pilot production is going to happen pretty close by but not within the same building. So what we have right now is an established baseline upon which we can run machine learning, training data based on the lab metrology experiments we've already run, and the designers of the ML technology are sitting right next to the chemical engineers and battery engineers themselves in order to continuously refine based on hypotheses in a way that's not usually done in private industry today. When we think about who our competitors are, we put them into three buckets. People who do R&D, people who apply machine learning to physics based models, and people who industrialize and scale cathodes. Nobody does all three. To be clear, we don't do all three yet. We don't do the third category yet. Industrialization is our next step. But even combining ML and R&D in the same roof through this approach. Okay, Will Chu and Churu Gopal, my two co-founders, can come here and talk about this slide for just an hour. So maybe we'll do that at a future energy seminar, but I hope that answered your question. The expertise is public source knowledge. Just search on Google Nature, Will Chu, and you will find the paper that is the basis upon which all of this has been built. Now the secret sauce of how we apply this to go from today's LFP to the sweet spot, that is our still internal advantage and trade secret that we're not going to reveal yet. But the application of machine learning to speed up R&D cycles is something that we've seen in other great platform industries. This is the way that EDAs work in semiconductors. It's the way that accelerated drug discovery worked in biotech. This concept is not new. The application is new. On that, you can see in previous energy seminars by Will Chu and his team, little hints, little glances of how this gets in. We also have, I don't know if you've met Austin Sendem, who was actually having been a Stanford Energy Club. He was the only student other than Divas to actually do an energy seminar. You can watch his video. That will give you a little bit of an idea of the space that they're operating without keeping away from the secrets. Any other questions? We have just a few minutes left. You can't see it, but I'm smiling through this mask. So I hear that you're wanting to create these highly specialized half of the ability to continue to make the old ones. So in other words, will my old vehicles, as they're battered in vehicles, will still continue to replace their batteries in their older vehicles? The way in which we should think about the electric vehicle as a system is disaggregating the body of the vehicle from the battery pack itself. The body, the way in which automotive engineering is shifting in electric future is that the body will be agnostic of the battery pack itself. And so by no means is it that, you know, I'm just going to throw an example, a 2015 Tesla Model S can only use a certain type of battery cell for the 2015 Model S. If you take a 2015 Model S and you take a 2022 Model S, the battery cells are actually very different from each other. They're just not visible to the naked eye. Great. Well, I look forward to catching up with the students who are joining me for the Q&A after this. So with that, let's thank Givas for an inspiration on the seminar. I want to hold you to that two-year plan. Some of you will still be there, so please come and celebrate the success of his company. Thank you very much. Let me turn off the mic.