SES AI Corp Q4 FY2024 Earnings Call

Good afternoon. Thank you for attending today's SES AI Fourth Quarter 2024 Business and Financial Results. My name is Shana, and I'll be your moderator for today. I'd now like to turn the conference over to our host, Kyle Pilkington. You may proceed. Hello, everyone, and welcome to our conference call covering our fourth quarter and full year 2024 results. Joining me today are Qichao Hu, Founder and Chief Executive Officer, and Jing Nealis, Chief Financial Officer. We issued our shareholder letter just after 4:00 p.m. today, which provides a business update as well as our financial results. You'll find a press release with a link to our shareholder letter in today's conference call webcast in the Investor Relations section of our website at ses.ai. Before we get started, this is a reminder that the discussion today may contain forward-looking information or forward-looking statements within the meaning of applicable securities legislation. These statements are based on our predictions and expectations as of today. Such statements involve certain risks, assumptions and uncertainties, which may cause our actual or future results and performance to be materially different from those expressed or implied in these statements. The risks and uncertainties that could cause our results to differ materially from our current expectations include, but are not limited to, those detailed in our latest earnings release and in our SEC filings. This afternoon, we will review our business as well as the results for the quarter. With that, I'll pass it over to Qichao.

Thanks, Kyle. Good afternoon, and thank you for joining us on our fourth quarter and full year 2024 earnings call. Today, we are pleased to present an update on the new era for SES AI as we reap the benefits of our All-in on AI strategy, deepen existing EV relationships, move to the next stage of our evolution to a revenue-generating company and share our plans for additional revenue streams with new market expansion in drone/robotics and Battery Energy Storage Systems, BESS. As you may have seen, we have made several announcements over the last several weeks, including that SES AI is reporting revenue for the first time in the fourth quarter. We are no longer a pre-revenue company and look forward to continuing on our commercialization path by growing revenue in 2025 and beyond. First, I would like to address the highlights in our EV space. Since the fourth quarter, we signed contracts totaling up to $10 million to develop AI-enhanced lithium-metal and lithium-ion batteries for EVs with two automotive OEM partners. These contracts solidify our relationships with our OEM partners and continue to demonstrate the value these partners see in the use of AI for battery material development, especially the value that an enhanced electrolyte can provide for both lithium-metal and lithium-ion batteries. Second, I want to address the important work we have accomplished in Urban Air Mobility or UAM drones and robotics. In January, at the CES 2025 show, we unveiled an AI-enhanced 2170 cylindrical cell for humanoid robotics. Earlier in 2024, we signed a significant purchase order with Data Blanket for drones for forest fire management and with SoftBank for HAPS communication satellites. This was a game changer as these AI-enhanced 2170 cells are the first batteries to use an electrolyte discovered by SES AI's Molecular Universe effort, which maps the physical and chemical properties of the entire universe of small molecules suitable for battery electrolytes 10 to 11 molecules. Our human scientists and AI agent learned that there's great similarity between lithium-metal and high-silicon lithium-ion cells in terms of electrolyte anode interfaces, and our development in lithium-metal electrolyte could also address high-silicon lithium-ion challenges, especially swelling and cycle life. We are now in both lithium metal and high-silicon lithium ion. 2170 is an industry standard format and is available on a large commercial scale. The electrode discovered through our Molecular Universe effort is a drop-in replacement. This allows us to have a CapEx-light model through contract manufacturing and to deliver a high-energy density and high-power density cell for customers in humanoid robots, drones, EVs, power tools and other applications. Third, SES AI is planning to expand into a new and growing field, Battery Energy Storage Systems, BESS. Last month, we announced that SES AI signed an MOU with AISPEX targeting up to $45 million to provide up to 100-megawatt hours of advanced Battery Energy Storage Systems, BESS solution with AI for Safety and the first deployment at a crypto mining site in Texas. This relationship is the first of its kind for SES AI, and we believe that this is something that can be replicated with additional partners in the future. This project will involve our battery safety platform, Avatar, which includes our AI for Manufacturing and AI for Safety solutions. Powering our commercial success are our two focus areas: upstream material discovery, Molecular Universe; and downstream battery safety, Avatar. In Molecular Universe, our collaboration with Nvidia on GPU-enhanced computational chemistry software helped accelerate this effort to map the entire 10 to 11 universal small molecules from more than 8,000 years to just a few months. We have now mapped 10 to 8 0.1% of the Molecular Universe. We think this is the world's largest single-molecule density function theory database, and we're also mapping solution and interface levels. This unique proprietary synthetic data helps train our AI model that will accelerate novel electrolyte discovery, as we have seen for lithium metal and high silicon lithium ion and later, even LFP cells for BESS and other chemistries. With Avatar, we have developed a new foundation model to monitor battery health and predict incidents before they happen. This AI-based model works in conjunction with a conventional physics-based battery management system, BMS, but is agnostic to chemistry, form factor emission profile and currently has been pretrained on manufacturing data and cycling data across a variety of cell chemistries, including NCM, LFP cathode and high-silicon graphite and lithium-metal anode and form factors, including pouch, prismatic and cylindrical emission profiles ranging from BESS to EV, drones and robots. In closing, SES AI has seen some exciting advancements during the end of 2024 and at the start of 2025, especially with us recognizing revenue for the first time in our history. All-in on AI is not only critical but is the future. We see great things continuing going forward. Here are some of the things I see happening: the change in our business model. Originally, we envisioned CapEx-heavy plans to manufacture cells for EVs and UAM at scale and then offsetting some of the required costs through JVs. Now our business model focuses heavily on selling our AI models and core battery materials, contract manufacturing of cells using our electrolyte and selling of contract imported cells. A focus on hiring of AI scientists and sales and marketing teams to pursue greater expansion of revenue opportunities from 2025 to 2027, continued evolution from a pre-revenue, EV-focused battery technology company levered to one battery chemistry to one that can leverage its prior work, along with its AI technology and AI-enabled electrolytes to generate revenue in a CapEx-light model much earlier than anticipated and with an ability to tap into two new adjacent verticals that we believe will be much larger than EV. Thank you for your continued interest in SES AI. And now I'll turn it over to Jing for financial updates.

Good afternoon, everyone. I will discuss our financial performance for the fourth quarter and full year 2024. I am pleased to share our financial results and provide insights into our fiscal health, strategic priorities, commercialization plans and 2025 outlook. For the fourth quarter ended December 31, 2024, we reported Q4 2024 revenue of $2 million, marking the beginning of our evolution into a commercialized and revenue-generating business. This revenue was primarily driven by contracts with our automotive OEM partners to develop AI-enhanced lithium-metal and lithium-ion batteries for EV applications and initial delivery of our lithium-metal cells to our customers like SoftBank for HAPS and Data Blanket for drone applications. Importantly, this revenue comes with a 63% gross margin, demonstrating the strong value proposition of our technology and All-in on AI strategy. Our GAAP operating expenses were $30.4 million for the quarter, primarily driven by research and development initiatives and general administrative costs. We utilized $12.3 million in cash from operations and invested $0.2 million in capital expenditures during the quarter. We concluded the quarter with a strong liquidity position of $262.5 million, ensuring our ability to fund ongoing and future projects. For the full year 2024, we achieved a total cash usage in operations and capital expenditures of $78.3 million, below the low end of our previous guidance of $80 million to $95 million. This includes operational cash usage of $66.1 million and capital expenditures of $12.2 million. As we look ahead, SES AI is well positioned for continued growth and disciplined investment in 2025. The greater visibility we now have on revenue has allowed us to provide some guidance for 2025. As you know, this will be the first time that we are providing this outlook. Therefore, we are being conservative in our expectations and focusing on the full year outlook rather than any particular quarterly cadence of revenue or expenses. We anticipate 2025 revenue to be between $15 million and $25 million, reflecting growth from our expanding partnerships and commercialization efforts. We expect to generate revenue from the development of AI-enhanced lithium-metal and lithium-ion batteries for EV applications, delivery of battery cells to UAM, drones and robotics customers and delivery of integrated Battery Energy Storage Systems with our AI solutions. We forecast total cash usage in operations and CapEx to be between $70 million and $80 million, prioritizing R&D advancements, commercialization and strategic market expansion with a capital-light business model. Our strong liquidity position extends our runway and continues to support our long-term vision with a focus on financial discipline and strategic investments. With our CapEx-light business model, we expect to maintain adequate liquidity well into 2028. We see tremendous organic and inorganic growth opportunities ahead of us. 2024 was a transformative year for SES AI. Achieving our first revenue milestone with a positive gross margin is a pivotal moment that signals our evolution from an R&D-focused company to a commercial and revenue-generating business. With high-margin revenue, growing partnerships and disciplined financial management, we are excited about the opportunities ahead. We appreciate your continued support and confidence in SES AI. Thank you. Now I will turn the call back to the operator.

Hi. Thanks for taking my question, Qichao. I was wondering if you can talk about the shift in strategy here. In particular, what led to this shift? And then it was mentioned that there were signing of contracts totaling up to $10 million for two leading auto OEM partners. If you are able to reveal, are they the same OEM partners that SES has JDA partners with? And how is it different from those previous ones with Honda, Hyundai? Thanks so much.

Yes, Winnie. Initially, we concentrated on lithium metal for electric vehicles, but our work extended beyond just battery cells to fundamental material discovery for electrolytes and safety in lithium-metal batteries. We realized, along with the OEMs we partner with, that these advancements could also benefit lithium-ion technology. By expanding our focus to multiple battery technologies and chemistries, we significantly increased our potential market size. It was a clear and exciting transition from a single technology in one market to multiple platforms across various battery technologies. Regarding the two OEMs, you can likely guess they are the ones we have been collaborating with for some time. They recognized the applicability of our innovations from a single chemistry to various battery chemistries across multiple markets, not just electric vehicles but also in drones, robotics, and urban air mobility. This shift in strategy was driven by the opportunity to address multiple markets.

Got it. Thanks so much. I just have a follow-up on the contract manufacturing. Are you able to unveil any sort of economics behind that? It's interesting that the gross margin was 63% in Q4, which was very impressive. Is this sort of like an initial type of metric that we should look at? And how will this evolve, I guess, heading into 2025?

Yes. I think it's too early to address that. And that margin was actually a mix of the AI models, the service as well as the actual selling of the cells. But going forward, in terms of the contract manufacturing, our goal to do contract manufacturing is to have a high margin. And I think it's well known that these next-generation battery companies just cannot compete at scale with any of the large incumbent battery players. So that's why we do contract manufacturing to achieve high margin. But the actual numbers, that's going to evolve based on volume, based on the particular application and customers.

Hi, everyone. First of all, congratulations on achieving revenue. It's a really significant event for the company. I did want to follow up on the gross margins, though. It seems like there are two different ways that you guys are sort of transitioning the business model. One is kind of, let's say, selling software versus selling hardware. Could you talk a little bit about how you see both the revenue opportunity for that and how that could potentially impact margins going forward?

Yes. Obviously, we want to sell just the software because that's a pure high-margin business. At the same time, for both the materials and the battery safety, it's hard to just sell the software without the hardware. So the hardware is like a hook to get a customer and then you sell the software on top of that. And the hardware on the material side, we contract manufacture these cells. And the cells are like a carrier for our materials, and that demonstrates the benefit and the power of this Molecular Universe AlphaSense software. So initially, the margin is going to be a mix between the sales and the software. But once we grow this demand, we plan to shift to more of a pure software business. And the same thing for Avatar. The pure Avatar itself is, of course, very high margin, but then it's hard to sell just software to BESS. You want to attach it to hardware. And that's why initially, for example, in the partnership with AISPEX, the entire solution involves the BESS hardware with the software.

So that definitely makes sense. Should we expect then going forward that the company will be able to layer on something like ongoing subscription revenue as it's selling these batteries?

Yes. That is the plan. And then we have a few discussions with a few battery companies where we actually offer subscription to the models. For example, that map that we show in the shareholder letter, that is a treasure map that tells any battery company or any company that needs to develop a battery where to look for in terms of new electrolytes. So that map, we sell products discovered from that map, and we'll also sell this AI model that was used to create that map. That's the annual subscription that you mentioned.

Hi Qichao. You have Mark Shooter on for Jed Dorsheimer. I'll echo those sentiments. Congratulations again on the quarter and the revenue outlook. The business is fundamentally different with multiple different revenue streams. So in that guide for the revenue, can you give a rough breakdown of where you think that's coming from?

Yes. So this year, just roughly high level, as we mentioned in the letter, we expect the bulk of this year to come from BESS and the rest to come from a combination of the EV and drones and robotics just this year. Now next year, and then going forward, that could change.

Got it. Okay. Regarding the BESS application, are you providing the complete solution with contracted manufacturing facilities as well as pack integration and the battery management system?

Yes.

Okay. And is that a difficult market to break into?

Good question. It depends on the market. It's a very fragmented market. I think you have behind-the-meter and also in front-of-the-meter and then some parts of the market are obviously pretty dominated by the big players and some parts are more fragmented. But at the core, in terms of the types of cells they use are actually quite common. You see about four to five different types of cells being used across all these different parts of the market. By doing that, we gain access to all the data. So the manufacturing data, the formation data, these data we use to pretrain the Avatar, and then we use the life cycling data from these to train the Avatar to monitor health. The nice thing is these data come from the same cells utilized by the larger players. Once we break into the medium size and get the data, we can train this Avatar, and then it can be used as a demo to show the larger players.

That's very helpful. And last one for me regarding the drone applications. Where are your contract manufacturers? And are any of your drone customers supply chain sensitive regarding the region?

Yes. For now, they are mainly in Asia. And yes, the drone contract manufacturers are sensitive and they would like us to move to a more U.S.-based manufacturing. We do have plans to do that. We have a phase where we go through this Asia-based contract manufacturing first and then switch to a U.S.-based.

Thanks. As with past quarters, we solicited questions from investors prior to the call. We received a number of submissions, and we'll go through a selection of those submissions now. The first question is: the performance limit today of cells with high-silicon anodes is the stability of the electrolyte. With SES' new electrolytes and the new advanced silicon anodes, what is the expected specific energy density, cycling durability and charge rates for 2170 cells made with these materials? What is the likelihood of these cells would be preferred for EV applications?

Yes. The high-silicon anode lithium ion behaves actually quite similar to lithium metal. In fact, the higher the silicon, the more it behaves like lithium metal from the perspective of this electrolyte anode interface. High silicon currently suffers from swelling and gassing challenges, leading to short cycle life. With this new electrolyte, we can significantly enhance the cycle life. The improvement in cycle life really depends on the amount of silicon that's used. If we go to 100% silicon, we actually expect to improve the cycle life. We expect to more than double the cycle life. The improvement is higher as we increase the amount of silicon.

Great. The next question we have is: can competitors copy the new solvent your AI discovered? Is it patentable or just trade secrets?

Definitely can. In this industry, especially if you have something that works really well, people will copy. The only way to compete is speed, and we try to focus more on the speed of innovation rather than just the innovation itself. For example, that map we showed in the shareholder letter is a very good example. We're not just finding a new electrolyte; if it works well, it will get copied. But in our case, we're not just selling a new solvent. Each dot on the map is a solvent. We're selling the map, and this map, no one else has that, along with the tool to create this map to find this map, no one else has that. So we have the ability to know where these new solvents are for different applications. That ability, for now, no one can copy that.

Thank you. Going down the list, the next question: as AI technology evolves, how is SES improving its AI models to enhance battery performance, safety and manufacturing efficiency?

The cost of AI has come down a lot, which has allowed us to train these models, both large language models and foundation models with more data. We are getting more data and both actual data from battery recycling and also a lot of synthetic data, basically data that did not exist before, especially for the molecule properties. The more data we have and the more data we use to train these models, the more sophisticated they get. The costs and efficiency have improved dramatically. It’s now possible to run training on these massive databases in a much more cost-effective way than before.

Thank you. We'll do one more from the pre-submitted questions. EVs are the number one market for 2170 cells. If SES has a better one, what is the prognosis for selling these into automotive? It seems the high-silicon version might be an earlier candidate than lithium-metal cells.

So the high silicon and lithium metal are sort of parallel tracks. Previously, we focused just on lithium metal. Now we are improving the cycle life of high silicon. We have been working with three OEMs closely. A few quarters earlier, we mentioned that there were OEMs that used 2170 cylindrical cells that we did not address because we only made pouch. Now, our pouch lithium metal is addressing OEMs that use pouch, and our 2170 cylindrical high silicon is for OEMs we did not address because we didn't have the form factor, allowing us to tap into a significantly larger target market.

Thank you. With that, I'll turn it back over to the moderator for a conclusion of the call.

Our first question comes from Winnie Dong with Deutsche Bank. Winnie, your line is now open. That will conclude today's conference call. Thank you for your participation, and enjoy the rest of your day.