Welcome to XtalPi's 2024 annual results announcement. The participants are currently on mute. We will now announce the housekeeping rules. This event is for invited investors only. The audio and text are for internal use and must not be publicly released. XtalPi has not authorized any media to distribute content. Unauthorized reproductions are considered infringement, and XtalPi reserves the right to pursue legal action. XtalPi bears no responsibility for any losses or liabilities arising from unauthorized distributions. Please be aware that investment carries risks, and make investment decisions carefully. Before we start, I would like to kindly remind you that there will be a Q&A session after the main speakers. We will start officially.
Thank you. Good afternoon. I am Gong Xinyang, the Investor Relations Director from XtalPi. Welcome to our 2024 annual results announcements.
Joining us today are Dr. Wen Shuhao, Co-Founder and Chairman of the Board; Dr. Ma Jian, Co-Founder and the CEO; and Mr. Tan Wenkang, CFO. Dr. Wen and Dr. Ma will introduce our strategies and business development. Mr. Tan Wenkang will present the financials. First, let's give the floor to Dr. Wen Shuhao, Co-Founder and Chairman of the Board, please. Thank you.
Thank you, Xinyang. Welcome to XtalPi's 2024 annual results announcements. This is also our first annual results since listing. We are very honored to announce our annual results for the very first time. In 2024, we actually achieved CNY 226 million. We achieved a 73% of the increases. The growth is getting faster, which shows that we are already in the new stage of implementation, if not explosion. After adjustment, net losses, it was CNY 475 million.
We're also very glad to see now we already achieved the threshold for the Hong Kong Stock Exchange, which also demonstrates with our joint efforts, the company right now is steadily striding into the scalable development stage. In terms of operational efficiency, through AI plus high-efficiency operations, we've achieved great results. The monthly average cash consumption dropped by 23% to CNY 48 million. The high growth and steady development will support our long-term vision, which any company must have. What is XtalPi's long-term vision? That is to create vertical artificial super intelligence or ASI in life sciences and future materials. What is vertical ASI? It is more accurate than general AI. While general AI might still experience AI hallucinations, vertical ASI must be precise and capable of genuinely designing effective drugs and materials. How to achieve that. The first step is correct digitalization.
Quantum physics and quantum dynamics are the foundational principles for understanding the molecules in life sciences and future materials, for ensuring model accuracy. Precise data on how drug molecules interact with the human body, data on how material molecules stably exist in devices and exhibit different physical and optical electronic properties are all understood through the first principles of quantum physics and dynamics. Data from quantum science is critical for the accuracy of vertical AI models, is quintessential, a fundamental basis for XtalPi's models. The molecules imagined by AI must also be synthesized in the real world. Previously, experiments were done by humans. Now we find that experiments done by AI-trained robots are more accurate and reproducible. They are supplemented with data from more positive and negative samples, which is essential for the synthetic efficiency of the molecules. That is the second correct industry digitalization.
With these two data sources, we move towards AI. The implementation of vertical AI is divided into two levels. The first is general AI, open source programs such as DeepSeek, which despite its room for improvement in accuracy, is growing with more users and stronger performance. We can stand on its shoulders. Meanwhile, with data derived from first principles in quantum physics and dynamics, and good data from robot experiments in specific industries, both data plus general AI can create our vertical ASI, which can be applied in life sciences and future materials, which the government and the nation really attach great importance to. With this accuracy and real-world experience, can be used to develop most valuable drugs and materials, and that is our long-term vision. When DeepSeek breaks down the barriers to algorithms and computing power, the applications of AI will become the future direction.
Where are the killer applications similar to those trillion-dollar level giants in searches, e-commerce, and socials? Well, in the era of AI, besides text-to-video, text-to-image, and answering questions, accurate real-world vertical ASI for developing high-value drugs and materials would definitely be a trillion-dollar application scenario. Vertical ASIs will assist mankind to develop microscopic molecules and materials with the highest added value more efficiently and accurately in a larger chemical universe. It could be drug molecules with NUCs in the billions of dollars, super materials to solve energy, collagen, climate challenges, or popular consumer and cosmetic product molecules. Currently in this field, given its team size, number of algorithm models, industry patents on it, or the scale of the robotic experiments, or just computing power itself, XtalPi is globally leading and one of the global leaders. We have nearly 300 PhDs covering fundamental sciences, AI, robotics, and experts in drugs materials.
With the recent introduction of specialist business professionals, such a brilliant pool of talents, 100 scientist robots are currently operating autonomously. Additionally, we adhere to business orientation without aimless expansion. We have raised the highest funds in our field. We had recently two placements. Now in terms of cash reserve, we have exceeded HKD 6 billion. We are the only publicly listed company in the sector in the Asia-Pacific region, with the Hong Kong Stock providing fundraising flexibility. With our capital strength, we will continue to maintain our leading position. In the capital market, XtalPi has consistently maintained very good momentum. On June 13th last year, XtalPi became the first stock listed on the Hong Kong Stock Exchange's Chapter 18C for hard technology, marking a watershed and milestone event for the Hong Kong capital market.
Our position also is translated into good performance of our shares, our stock, and its daily trading volumes reach HKD 2 billion-HKD 3 billion at peak times. We have received lots of funding and our strong liquidity success led to two rapid placements in less than two months, raising over HKD 3 billion. Following each placement, our market value has continued to rise. This trend has established an efficient fundraising channel for company, allowing us to secure more funds to expand our business into more industries and collect more industry data, all of which will support XtalPi's long-term vision of the building super AI for industry verticals. XtalPi's global influence is rare and unique.
In AI drug development, we are one of the very few companies that has long-term collaboration and repeat purchases with and from top pharmaceutical companies in Europe and America, such as the J&J, Eli Lilly, and Pfizer, with whom we have 10-year long strategic cooperation for the targeted drugs development. For Goldman Sachs China AI Medical Care Index, it includes us as a primary rated stock. Cathie, with ARK Invest report on significant innovations in 2025, mentions us as the only company from the Asia-Pacific region. The WAIC, XtalPi, along with Huawei, Qualcomm, and Comac, received the highest award for excellence in AI, Super AI Leader. Our foray into material industry has been covered as a front page story by Forbes and MIT. Our global influence and market capabilities make us one of the most unique AI companies in this region.
We've always managed to partner with top global players in pharmaceutical, technology, and research. That is because of our solid technologies and market strategies. I just mentioned, we have a 10-year strategic cooperation agreement with Pfizer and a $250 million partnership with Eli Lilly for a single target. We co-established a future lab with MIT and signed a strategic partnership with Microsoft. In materials, we also collaborate with the industry leaders. Recently, we've reached a cooperation with the government of China, as well as the royal family in the Middle East. For example, in China, with the Ministry of Industry and Information Technology, we've created an innovation and development enabling platform for SMEs. This is the first of its kind. It will empower SMEs across various sectors with a focus on niche industries like beauty, consumer products, pharmaceuticals, and cutting-edge new materials and agrotech.
We cooperate now with U.A.E. royal family to empower traditional pharma research in the Middle East with $30 million. We also partner with Indonesia's leading conglomerate, Sinar Mas, to advance AI industry in the Asia-Pacific region. We also collaborated with Hengdian Holdings, the largest data investment platform in Guangzhou, for an innovation alliance for AI tech in the Greater Bay Area. What we really want to show you as the management team is what we mentioned in the long-term vision, that is the source of our data. That would be the data engines and evolution engines for building vertical industry super AI.
This includes clusters of robots used for Traditional Chinese Medicine extraction and separation in our labs in Guangdong, Hengqin, where the general secretary spent the longest time during his visit for Macau's return celebrations last year. In addition, we also have robotic systems developed for Amgen, the U.S. pharmaceutical company, to create compound libraries and scientist robots for catalyst screening used by Academician Yang Weiming and Sinopec Shanghai Research Institute of Petrochemical Technology, as well as robots for electrolyte development assisting Academician Jiang Qin, the Vice President of Fudan University, and systems for chemical reaction research with Academician Masaaki at Fudan University. Unlike robots in other industries, XtalPi's high flexibility, high precision science robots evolving in academic laboratories are specifically designed for drug material research. They're characterized by high data quality, execution efficiency, and it can continuously generate high-quality data 24/7.
Their data efficiency is 40x higher with a 90% accuracy rate in specific fields, while continuously accumulating data for further evolution. The precision evolution of individual scientist robots lead to the clustering and scaling of industrial robots. The Greater Bay Area boasts unique advantages in hardware, robotic industry chains, supply chains, and cost. XtalPi's presence in the Greater Bay Area allows us to leverage those resources, achieving initial scalability with hundreds of robots operating autonomously. XtalPi will maintain our lead and become one of the first companies with thousands or even tens of thousands of such scientist robots. This will facilitate our business and data scaling, thereby creating our vertical ASI for the industry, namely our vision.
The videos they're looking at demonstrate the need to engage with hundreds of industrial chain suppliers in the Pearl River Delta and the Yangtze River Delta to build such precise scientist robot systems that interact seamlessly with major clients in Europe and America, which is also a current advantage of China. These sections showcase the various types of molecules our systems have handled and developed, ranging from small molecules to peptides and antibody drugs, as well as various new materials, including perovskites and catalysts. In terms of pharmaceuticals, we are involved with the COVID-19 oral small molecule, PAXLOVID, which is already on the market. We also cleverly published articles with Pfizer, which have already been translated into significant commercial and social value. This is another case we would like to showcase. We can call that future drug with AI optimization. Call it future medicines.
This is based on AI precision regulation, which optimizes dosage. It is based on a deep understanding of biological mechanisms of metabolism-enhanced immunity. The first cohort of patients achieved 100% of complete remission rate with extremely low dosage of 0.1%, eliminating life-threatening cytokine storms at 1/10 of the cost. This is the optimization dosage that is achieved in the CAR-T therapy that we're looking at. It also dramatically reduces the cost of production, as I said, to 1/10 of the current cost. That's how AI is creating future medicines. It's effective, safe, comprehensive, and low cost. It's affordable for the general public. There is a case of a lymphoma in a 69-year-old paralyzed elderly with lymphoma that had invaded the central nervous system and has regained ability to stand.
Acute leukemia, which is most harmful to children in China due to environmental exposure to formaldehyde. Previously treated primarily with radiotherapy and chemotherapy, which would often result in disabilities. Now with AI optimization based on targeted therapy and dosage, now they have hope. The first cohort of children, as you can see in the photo, are not only fully cured, but they are returned to the PE classes. With optimization, the prices also reduced tremendously so people can afford it. This is the actual significance of AI. Equal rights enabled by technology. We published papers on "Nature." The potential value exceeds $100 billion. I will stop here. Next, Dr. Ma Jian will present our business landscape. Thank you.
Okay. Thank you, Shuhao. Hello, everyone. I am the CEO, Ma Jian, of the company. Let me update you about our business progress last year. Last year, we continued to deepen the technological capabilities of AI plus robotics and change the R&D paradigm from the traditional way of letting humans design molecules and design and perform experiments, to the new paradigm where AI design molecules and experiments and robots perform the experiments. Thus, with this paradigm shift, we can open up the era of intelligent drug and the material R&D. We have built a unique flying wheel of a high throughput experiment, high-quality data, and intelligent model, and empowered the R&D personnel in our daily operations, accelerated delivery process that broke the bottlenecks in drug discovery and R&D of new materials molecules, and significantly expand the explorable space of chemistry.
We believe that in verticals, AI-first science model driven by high-quality data as its core will become a disruptive force. In 2024, we built another 200+ existing models and made more significant breakthroughs. In terms of data, our robotics lab has covered more than 80% of the common pharmacological reaction types and is operating 24/7, generating high-quality data assets with ultra-high throughput at all times. Each month, we can accumulate over 200,000 reaction process and data, and the data collection efficiency is 40x that of the traditional method. The quality of these data is far better than the open-source data or the manual experiment data. Thus, our AI model can obtain very high accuracy and confidence. In terms of models, we have built more than 20 kinds of AI reactivity and experimental condition prediction models in 2024. Their accuracy all exceed 80%.
AI models can predict successful and failure responses better than experimental experts. Particularly, the prediction accuracy of failure responses is more than twice of the experimental experts. In addition, based on the first principles of quantum physics and the data set generated by mobile laboratory, we built our own UV spectrogram prediction models and LC-MS spectrogram-based yield prediction model, which can obtain experimental yield without doing product separation and purification, and accuracy is over 90%, which made our data labeling far more efficient. We also autonomously develop multi-agents to build intelligence and automation of whole process from molecular design synthesis to post-processing, molecular detection, and data analysis in order to improve the experimental efficiency, reduce experimental threshold, and complete the all-scenario coverage of software and hardware.
For example, in the synthesis stage, we use agent to analyze the spectrum of the reaction process, determine the reaction structure, recommend the separation method, and judge the quality inspection result, and push the reaction process. In the management process, we use agent to ensure the accuracy of the process data and consolidate product reports, and finally, deliver the molecules. With these capabilities, we can efficiently conduct chemical experiments and explorations and accumulate large-scale, high-quality experimental data to feed AI algorithm and model. In the field of drug discovery, the pipeline we have empowered also hit numerous milestones in 2024. In 2024, we collaborated with multiple leading biotech companies to efficiently design and discover lead compounds for a number of difficult targets, which were highly appraised by the partners and technology committees.
In addition, following the successful development of an innovative drug development project with a leading bio-pharma, we conducted several new drug development collaborations this year. We also plan to further expand the scope of cooperation in 2025 to jointly tackle more targets and indications. In particular, the pipeline we serve has entered a phase I of clinical trials for the first time. The global-first targeted candidate drug for treating diffuse gastric cancer developed in collaboration with Signet Therapeutics, received IND clinical trial approval from both FDA and NMPA in June last year. The first administration to a solid tumor patient was successfully completed in Beijing Cancer Hospital. This pipeline received Orphan Drug Designation for gastric cancer from FDA and the Fast Track designation from FDA in February 2025. This is also the world's first case of organoids plus AI-enabled innovative drug design and screening.
Our drug candidate for primary hyperoxaluria developed in collaboration with Merida Biosciences has also received a Rare Pediatric Disease Designation from FDA and expected to address the dilemma of the drug availability for patients with type 2 and type 3 hyperoxaluria in the future. The 25-year pipeline is scheduled to enter clinical phase I. The American Big Pharma, with which we previously signed a $250 million partnership agreement, is well on its way and our partnership expanding in 2024 and onwards. The XFF force field model we co-developed with Pfizer was published in academic journal, the two companies will continue to iterate on a new generation of force field models for Pfizer proprietary chemical space in 2025. We moreover also made progress in new areas such as cutting-edge therapies.
For example, in the field of CAR-T, Limo, the biotech company we incubated completed the first drug administration in patients with systemic lupus erythematosus. In addition, progress was made in tumor vaccines, AI polypeptide R&D platform, non-invasive ophthalmic drug delivery platform for nucleic acids and small molecule drugs and mRNA. In terms of small molecule business, we have built XtalFold, a structural modeling platform, XenProT AI platform, Xentient discriminative AI platform, et cetera, continue to advance the technology of deep humanization, high throughput humanization, pH-dependent transformation. In 2024, we're also in the leading rank to develop the prediction algorithm, XtalFold. We developed cooperation with more than 20 multinational big pharma companies and leading bio firms such as Johnson & Johnson and XtalPi.
We can also accurately predict antigen-antibody complex spatial structure within a day using only amino acid sequence information, thus lead the industry for success rate and the modeling ability in difficult model regions. XtalFold has been fully validated for more than 30 internal external projects and has achieved great results in scenarios such as antigen design, epitope recognition, affinity neutralization, pH sensitivity modification, and dual antibody design. The good data can be also generated from XtalFold and fed into our multimodal AI platform, reinforcing our cutting edge in terms of humanization. In 2024, we also expand our applications to the global market. In 2024, we won a bid for Guangdong Provincial Laboratory TCM New Drug Intelligent Automation Integration Innovation Platform Construction project, which has achieved phase delivery and confirmed acceptance. This is also the first set of automatic platform for separation analysis of the effective TCM ingredients in China.
In 2025, we also officially signed a commercial cooperation agreement with the U.A.E. royal family, including a first installment of $30 million cooperation, which will build U.A.E.'s first automatic modern R&D experimental platform for traditional medicine in the Middle East and create a new paradigm of traditional medicine R&D in the region. With decades of technology accumulation in the industry, we have a closer loop of AI plus robotics to help material discovery research. Based on the common underlying technology, we have built a database engine on top of it to empower our AI models for vertical applications. We have had a full stack technological capability encompassing domain knowledge, industry specific models, AI understanding of the problems, accumulating the right data, and building the R&D infrastructure.
Through the exploration of AI drug development, we have accumulated rich, in-depth knowledge related to molecules and crystals, which is equally important for the development of materials. We have built our own various models in molecular physics, quantum chemistry, molecular dynamics, crystal, et cetera, and extended the application of our AI capabilities from pharma industry to more material fields through the deep integration of data in the new fields. In 2024, we performed advanced in the seventh CSP blind test held by Cambridge Crystallographic Data Centre, or CCDC, which involved multiple material systems such as the photoelectric material molecules, biomedical molecules, and pesticide molecules. We demonstrated our global leading technological strengths in this competition.
We achieved a blind test prediction of the experimental crystal patterns for all seven target systems, which result to 100% accuracy and made us one of the very few top teams in the world that can accurately predict the relevant and the relative stability between the different systems. Frankly, we internally feel that the competition questions were too simple. In some advanced science and technology fields, the academic understanding of the forward-looking developments may be updated slower than the industry at the forefront. In 2024, we also expanded the technical capability of solid-state research to materials and partnered with companies for projects in electrochemical materials or polymer materials and ceramic oxide materials. Some of these projects have been delivered and earned income. AI is empowering myriad industries.
Our AI plus robotics are also empowering more and more verticals including drug R&D, chemical industry, new energy, new material, and agriculture, with more fresh pathways made along the way. There are numerous such cases. If you wish to hear more, this year, there will be more business opportunities than 2024. Indeed, the number is growing, and the market is active. Particularly, there is explosive growth of China's tech sector. We are standing at the doorstep of a new era. In the new areas of drug and material, we see a vast number of opportunities and broad prospects. I would like to hand over to our Chairman, Dr. Wen Shuhao, to take over and brief you on our progress in new areas and strategic plans for the future. Thank you.
Thank you, Ma Jian. As you can see, we have made substantial progress in the technologies and businesses based on our underlying philosophy and our unique roadmap. Our new industry layout is now taking shape. I will now elaborate further on the pharmaceutical sector, where we will continue to maintain our leading position and data advantage through repeat purchases from our top clients. Our strategy in the materials field is similar. We aim to collaborate with top players in specific material industries because they are the most understanding customers who also have the most industry data. For example, in chemical catalyst, we work with Sinopec, in perovskites, we are collaborating with CN Carbon. We partner with Fangda Carbon, and in solid-state batteries, we team up with Yunnano Materials, all industry leaders. In the future, we will establish deeper collaborative research partnerships with more leaders.
For example, Peking University and Massachusetts Institute of Technology, we are also investing in incubating teams and projects from MIT and UC Berkeley's national laboratories. That is on the material sector. We deploy rather fast, even exceeding our expectations. For future agriculture, we have laid out plans for AI and SEAs, also special materials for desert environment, special fertilizers, super fertilizers, and super crops, et cetera. Phillip Bio is this super fertilizers product that we have. In the consumer goods sector, we have co-established the innovation platform for MSCs with the Ministry of Industry and Information Technology, specifically to empower SMEs in the cosmetic sector to develop customer product molecules, therefore shaping our future consumer goods segment.
We will upgrade our organizational structure, transitioning from the original XtalPi's AI drug discovery and XtalPi's AI for science to a three self-reinforcing flywheel structure, which consists of 3 key BUs or business units. BU Future AI-powered Molecule covers future drugs, materials, agriculture, and super consumer goods, all generating services, business, and most importantly, data from the industries. Another is BU Intelligent Robots, and that is on robots for only scientific research and for the creation of various molecules. These robots will continually evolve their precision, explore chemical domains, developing life-saving drugs, and creating remarkable super material molecules. Its sub-departments include flexible, high-precision, ultra-concurrent scheduling, supply chain scaling, and medical robotics, all generating services, business, sales, and data as well.
These two BUs, the BU Intelligent Robots and the BU Future AI-powered Molecule, will keep producing a substantial amount of data which can drive the company's commercial scale and generate significant data. This is the ultimate super artificial intelligence that we create for the industry, and that is what we often say as our ultimate goal. Through these precise, rather than AI hallucinations, super AIs for the industry will want to solve the mankind's greatest challenges, including questions about life, the dimensions of civilization, energy, and others. There are two cases of substantial progress in future materials, one for desert remediation. AI is optimizing both single materials and a combination of over 30 materials across the 3 major categories. AI agents help to manage the optimal use of light, water, and materials, making the remediation of every acre of desert commercializable and scalable.
The core is that our effort on each acre can be converted into profits, that's the key of desert remediation. China, the Middle East have deserts, there are hundreds of millions of people that live with hunger. We talk about applications of AI, well, we want to work on universal values, we want to apply where social and economic values exist. These are some of the cases we can show you. Here is another case for a consumer goods sector. It demonstrates AI-designed peptide molecules, they exhibit exceptional transdermal absorption efficiency and optimize the microenvironment of hair follicles. It's very exquisitely designed and is supplemented by AI-designed small molecules that can activate hair follicle stem cells. This combination of AI-designed peptide into small molecules serve as cosmetic raw materials. This is a consumer goods.
It completely outperforms the leading drug molecule, minoxidil. As a consumer product molecule, it is expected to secure global trade and production licensing through NICI by early next month. It has the potential to become a global best seller, generating significant cash flow this year for the company. The capability to design peptide molecules itself with transdermal properties presents immense application opportunities in the cosmetic field, which is also the reason behind the platform that we have with MIIT, the Ministry of Industry and Information Technology, and XtalPi. The boundless potential for AI consumer goods will lead to shorter monetization cycles and impressive cash flow. That is all on the important restructuring of our organization. Now I'll give the floor to our CFO, Tan Wenkang , to present our 2024 financials.
Thank you, Shuhao. Hello, everyone. Next, I will elaborate on the company's financials in 2024, maybe last year. Overall, as the chairman said, in 2024, the company registered revenues of CNY 266 million, up 53%, adjusted a net loss RMB 457 million, down 13%. The monthly average cash burn, RMB 48 million, down 23%. In 2024, the company realized robust revenue growth, particularly in the second half of the year. Revenue growth accelerated and reached 73% year-over-year. Moreover, customer base continued to expand from 75 customers in 2021 to over 280 in 2024. We are actively exploring high-quality clients at home and abroad, empowered biopharma and new materials R&D with our AI plus robotics platform. In 2024, we continued to implement our business globalization strategy and actively expanded our global customer base.
Among the regions, the U.S. contributed at 30%, while countries such as Korea, Japan, Sweden, and the U.K. took up 13% of the total revenue. In 2024, revenue from our intelligent robotics solutions grew rapidly, up 88%. This included our state-of-the-art research services, automated synthesis services for pharmaceutical and agricultural companies, R&D solutions, and commercialization of material science. Particularly, the modernized TCM solutions we provided to Guangzhou University of Chinese Medicine have been delivered and accepted. Second, drug discovery solutions. Revenue grew by 18% year-over-year with breakthroughs in our macromolecular business gaining recognition from several leading international pharma companies where drug development capabilities continue to be validated. We expect more milestones in terms of revenue growth in 2025. In 2024, R&D expenses totaled CNY 480 million, that focused on the R&D people's remuneration, spending for cloud computing, and material technologies.
More importantly, we focused our R&D on key areas. In terms of the general administration expenses, it was CNY 418 million. If we exclude share-based or equity compensation expenses, listing expenses, which is one-off, as well as depreciation and amortization, our general administration expenses amounted to CNY 232 million. Our sales marketing expenses stood at CNY 71 million, primarily covering the remuneration of the commercial people and the ongoing expenditures related to expanding our domestic and international business. Our contract performance cost was CNY 143 million, showing significant improvement in delivery efficiency. Its share in revenue decreased sharply from 72% in 2023 to 54% in 2024. Thanks to a strong revenue growth and improved operational efficiency, our adjusted net loss was CNY 457 million, down 13% year-over-year. With our efficient operating strategy, the average amount annual cash burn was about CNY 580 million, down 23%.
Moreover, our cash on hand is adequate. Our cash balance exceeded CNY 3 billion at the end of 2024. As the chairman said, by the end of March, that number rose to CNY 6 billion.
Thank you, Mr. Tan. We will stop here for the financials and the presentation of the management. We will now start the Q&A session. Thank you very much. Now we will officially open the floor for questions. Please raise your hand or leave the text for questions. Please press number one followed by the pound key if you were to join the Q&A session, or you are more than welcome to ask your question directly after unmuting yourself. Welcome. For those joining by telephone, please press number one followed by pound if you were to ask a question. For those joining on the app or on the internet, please unmute yourself and ask your question directly. nine nine two eight by phone. Please state your name and your organization before the question. Thank you.
Mr. Shuh ao, Mr. Gong , Mr. Ma. Good morning. I am Pan Ruoxian, and with CITIC.
Thank you for the opportunity. First of all, congratulations on the impressive performance in 2024. I just heard the very detailed presentation from the company, quite comprehensive already. I have one question. For materials science, we see that you are increasing investment. My question is, for your capabilities of AI and robotics, what is the underlying logic for you to expand into new areas? Thank you.
I can take the question first, and my colleagues can supplement later. Thank you. Thank you for the question from our analyst and investor. It is a very good question. Materials market itself is on a 10 trillion market we have to deploy. What is the reason for us to expand to that industry? Well, first, as I said, quantum physics and dynamics serve as our foundation when we understand the molecules of drugs and materials. It is the same.
Given two molecules in drugs or materials, they must exist in a stable manner, and the properties must show. These are the underlying theories and principles, and this is what XtalPi has strength at. My background is quantum physics, and we're trying to develop the life-saving drugs. The algorithms and underlying logic actually determine that algorithms are interchangeable. That's the dimension of the algorithm. For robotics and automation, it is the same. First of all, we think robots are more efficient at conducting experiments. When we design the experiments, we first had standard interfaces. When we graft our experience to new materials or other sectors, you would see the interaction between the receptor and the ligands are similar. The chemical experiments involve a lot of inorganic tests that require high temperature.
It's the same concept as drug discovery, and both require high flexibility and high crystallization design. That is why we can make such an expansion, or we can expand into that field, too. It has even better opportunities than the drugs industry. As Mr. Ma said, we have seen that when drugs molecules are designed, when they get to clinical trials, no matter how good the molecule is, there is a clinical trial stage. Materials actually have a very good cycle of iteration after materials are developed. They can be used on components and to see whether they can actually achieve automatic manufacturing and for more efficient material iteration. It represents a vast ocean of opportunities. Algorithms are on the underlying or the bottom layer, and we have the standardized interfaces with high precision and high flexibility.
It's the same whether you develop drugs or materials. That is also why from drugs, we can enter materials. From the perspective of the results orientation, we've already invested a lot, we said, from the early-stage discovery and also later investment. I'm not going to elaborate because there are requirements on NDA, non-disclosure. That's why we are in the materials industry, and we've seen so many progresses. Okay. Let's see whether Mr. Pan, his question is answered. Okay, thank you.
Thank you, Mr. Wen, for the answer. Very clear. Thank you.
All right. We will page through the next question. Thank you. one one one three on from the phone. Please state your name and organization before the question. Thank you.
Thank you. My name is Jeffries. First of all, congratulations on the great results in 2024.
I have a very simple question for the management for this year. We see that lots of attention of the investors are given to general models like DeepSeek. How do you really view the impact of these general large models on the company? Thank you.
Thank you, Jeffries. I can take the question. First, indeed, DeepSeek has brought a lot of impacts this year, and I would like to thank Liang Wenfeng, my Senior in Zhejiang University. I entered into the industry, and he's before me. Zhejiang University is very active, actually. Lots of sharing of the alumni from Zhejiang University. We stay tuned to recent technology advancement. For general models such as DeepSeek, it has received wide attention from the public, while many people want to know about the future direction of artificial intelligence.
For AGI, or artificial general intelligence, becoming more common, it can answer a lot of questions and write articles and program and write poems themselves for us. For specific scenarios of research and industry, there are often a lot of hallucinations, or it lies in plain terms. It fabricates articles and data and statistics. That's the trap of AI, which is based on the training models. In its essence, we lack high-quality data and materials for the training, so that AI can actually be used for vertical industry applications and for mathematics, theoretical physics, which will require a lot of logical reasoning.
It's really for math, AI is excellent, when we look at pharmaceuticals, drug discovery, materials, chemistry, et cetera, which require large amounts of experiments and subjects alike, we need hypothesis to be proposed. Then we need experiments in the real world to derive the real science data to verify the hypothesis or to modify the original hypothesis. This is a trial-and-error closed loop. Then we can supplement more solid and effective data to overcome the hallucinations, therefore AI can finally be applied in the science industry. Thus better empowering research and development and finding new materials and drugs. That is why when we look at the past five years at the development of AI itself, in 2016, AlphaGo was launched. Last year, AI was given the Nobel Prize for chemistry and also physics.
In this process, we've come to realize that data will become a focus as a future factor of production for algorithms, data, and computing power. In a lot of scenarios, computing power is no longer the bottleneck. If you don't need a GPT, like large language models, in a lot of applications, computing power is no longer the biggest bottleneck. For algorithms for DeepSeek and open-source algorithms, you don't need to repeat the programming. For reasoning, you can actually be based on large models such as DeepSeek. After these two factors of production, the whole focus is shifted to high-quality data, especially accumulation of digital data. Here at XtalPi, in the past few years, indeed, four or five years ago, we already clearly realized the future trend of technologies.
For AI application, it definitely needs to enable industries and even manufacturing sectors of growth. For our robotic laboratories, as we said earlier, AI can help mankind find new ideas. In general, AIs and AI agents can dismantle tasks. Robots can continuously work 24/7 in generating high-throughput data. The whole operational process is of higher consistency and reliability compared to humans. There are also negative samples that can be collected effectively. For with that, the AI technology platform, so AI plus AI robotic agent, can help us train vertical ASIs in the future and bring exponential gains in efficiency. That's how we can truly have the opportunity to create a disruptive super artificial intelligence, and that is our view on the future involvement.
Of course, most importantly, as I said, DeepSeek for the industry and also what we say on companies or even the societal development, it has injected lots of confidence. Lots of us feel that innovation and lots of time, the confidence comes from capital. In China, we also see that there are lots of technologies that skyrocket, which boost our confidence. For XtalPi, we also are dedicated to the cutting-edge frontiers of technologies. What we do is going to be globally competitive and leading in the future. Thank you.
I just want to quickly add, for DeepSeek for XtalPi is definitely good news. As Ma Jian mentioned, DeepSeek actually broke down the barriers of algorithms and computing power. Data become the core issue for XtalPi in the past. We've been in the industry for a decade.
Well, for biopharma, we had customer data for Pfizer and also Eli Lilly, and you see. Now we can actually bring down the cost, and we can also get to access the data that other peoples cannot. We can then generate efficient data effectively. If DeepSeek break down the barriers of algorithms and computing power, AI applications actually become the focal point in the future. We are the top leader in this industry, and we've already implemented molecules. Some of the molecules have already generated a great commercial and a societal value from these two dimensions. To answer your question, DeepSeek for XtalPi is definitely great news, as Ma Jian said, for the prosperity of the Hong Kong stock market, and it brings even longer benefits. Thank you.
Thank you for your question. The next one, please.
Okay. I want to thank the management for the answers. Next, I would like to invite the investor with a phone number ending two eight two zero. Please state your name and institution.
Okay. Good evening, members of the management. Thank you for taking my question. I'm Lu Wentao from CMBI. My question relates to international business. We have seen the company doing a lot of international expansions and implementation. Could the management kindly let us know how you will further advance your international market? What are the key regions that you are focusing on?
Okay, thank you for the question. Let me address it first. First, the company is scarce and rare because it is one of the few Asia-Pacific AI companies that can do global business.
We have always followed the principle that only by working global can we develop the biggest business. We represent the tech sector in China to work in the global market and do global business. We also captured opportunities from the big pharma in Europe and America. They are leading in this industry. In biopharma, we do a good job in securing the repurchase from those big pharma. That is key to biopharma. Another region we're paying close attention to is Middle East, which is rising and capital rich. Middle East needs technology very much. As a tech company in China, we have proven that we are able to help the leading European-American companies with the enhancement of R&D efficiency. In Middle East and the rest of the Middle East, we would like to expand our presence for the reason I just noted.
A big problem in the Middle East is the desertification and ecological challenges. We have made significant progress in this regard. Moreover, the Middle East is very keen on developing bio-pharma. The people lived in the Middle East, so they face the ecological challenges. Of course, they also want to live a safe and healthy way of life. We have very recently seen the $30 million deal for intelligent robots. We partner with the royal family to co-develop the traditional drugs in the Middle East. Another hot region is Southeast Asia, for which I would like to mention consumer goods. In Indonesia, for example, there are a lot of young people that are eager to consume. That is why we have partnered in depth with Sinar Mas in Indonesia. That is why we are so confident in our new organizational structure and the new business units.
We are so efficacious and efficient in developing drugs. We have such a constant technological advantages in China and the Middle East and Southeast Asia. We can address the local needs and the demographic trends. This is our strategy for the future. You can expect that in those geos, we are going to have significant business cooperation.
Let me add to what you said. Just now Shuhao emphasized that we are developing our presence in the Middle East and Southeast Asia for bio-pharma and in embodied AI , and also robot scientist. The leading European American big pharma are still our most loyal customer base.
Over the past year, in Shanghai, to which you're welcome to visit, you're welcome to come to our Shanghai and our Shenzhen headquarters, Eli Lilly, Merck, UCB, top management and other executives would invariably come to the company when they visited China over the past year. They would like to see the cutting-edge convergence of AI and robotics. Indeed, in China's industrial and supply chain and the technological base, we enjoy an advantageous position in China, and we are well-positioned to also lead the world. Before macromolecular business, we are achieving progress in business. Our team also did a global roadshow in the past month among the big pharma. For several intensive days, in-depth communication was made. In 2025, we are going to further advance our global layout, and we do look forward to excellent progress. Please do stay tuned.
Thank you, Mr. Ma.
Next, I would like to invite the next question due to time limit. The next question will be the last one, too. Thank you for the answers. I would like to now invite the investor with a phone number ending four three six eight. Please state your name and institution, please.
Thank you for this opportunity. I'm Cyrus from Deutsche Bank. I also congratulate the company on excellent performance. Recently, AI has been very active. Many companies are increasing the development in this regard. How will the company continue to maintain its competitive advantage in models and the data?
Thank you, Cyrus, for your question. This is also a question we have been also moving forward to access high-quality data, and on this basis, iterate our models so as to develop a data flywheel with which we will yield our competitive edge.
In summary, the continuous upgrading of the robotic scientist will also lead to the accumulation of high-quality data. Second, AGI has been better and better in terms of organizing the public data and patent literature. We are working on this tool. In the course of our business, we will also accumulate know-how in the various industries and sectors, which will be fed into our big models for domains. The engineer dividend in China and the scientist dividend in China will provide our capability in terms of integrating hardware and software and the associated cost advantages. More and more of our talents come from the top Chinese universities and the elite students there. By working with the institutions of higher learning and the research institute, we will access more of the cutting-edge knowledge and the best-of-kind talents.
All these are about our strategies for competitive edge in the future. Let me make two points of addition. Recently, we made two recent rounds of placing, which gave us abundant cash reserve. With that, we can do more business. When a company had not enough money, we would not partner with this company. Now, with more backing of capital, we can have a freer hand for partnerships so that we can access more industry data. Secondly, the government also provided valuable support. The Ministry of Industry and Information Technology has an SME center which told us that thousands of SMEs in China will need our support. This is evidence of the government support through the ministry. We can access thousands of the SMEs. We will be able to quickly access data there.
Third, if we have excellent team and data in the potential targets, we will execute our strategy of M&A so that we can benefit more from our capital and technological accumulation. That will keep us at the forefront and increase our competitive advantage and hurdle versus our competitors.
Okay, thank you, investors, analysts, and friends for your participation. Should you have further questions, please feel free to contact our IR team at any time. Thank you for joining us. That concludes this meeting, and I wish you a very good weekend. Thank you all