After more than 700 days, the youngest "Ai four little dragons" enterprise cloud has finally boarded the scientific innovation board from technology today! (Stock Code: 688327) at the opening of trading today, the share price of cloud technology rose 56% to 24 yuan / share. As of 10:20 p.m. of Zhidong, the share price fell back to the range of 21.06-21.8 yuan, and the latest market value was close to 16billion yuan.
▲ stock price trend of Yuncong Technology
Cloud technology was born in a laboratory of the Chinese Academy of Sciences. Zhou Xi, the founder, graduated from the University of science and technology of China and the University of Illinois at Urbana Champaign, and studied under Professor Thomas S. Huang, the "father of computer vision" in the United States. In 2011, Zhou Xi returned to the Chinese Academy of Sciences to establish an AI team, and in 2015 he founded Yuncong technology company.
▲ Zhou Xi, chairman and general manager of Yuncong Technology
It is mainly engaged in man-machine cooperative operation system and artificial intelligence solutions. Cloud mainly focuses on smart finance, smart governance, smart travel, smart commerce and other fields in terms of technology applications. In 2021, it has entered the camp of "1billion yuan" in revenue.
This time, the total amount of funds raised by cloud from technology is about 1.728 billion yuan. The successful listing means that it has more sufficient funds to invest in technology research and development and landing.
▲ the main purpose of this fundraising of cloud technology
So far, Yuncong and Shangtang have completed their IPOs in the "Ai four dragons", Kuangshi has entered the final stage of waiting for the registration results, and ETO technology has terminated the process. In addition, a number of computer vision players such as Yuntian lifeI, Geling Shentong, and innovation Qizhi have also been listed.
This is an epitome of Chinese AI enterprises moving towards a more mature stage. At the same time, it means that AI enterprises should put their business performance on the table and accept stricter market review.
The AI national team from the Chinese Academy of Sciences completed its IPO in more than 700 days
"We really want to change the world." Zhou Xi, chairman and general manager of Yuncong technology, said at the listing ceremony this morning, "today's AI is not only a black technology, but also our dream for the future and our determination to change the world."
This is an entrepreneurial story of post-80s doctors.
Born in Neijiang City, Sichuan Province in 1981, Zhou Xi studied at China University of science and technology from 1999 to 2006, and then went to the United States to study under Professor Thomas S. Huang, the "father of computer vision". In that year, Professor Huang enrolled only three students in the world. After studying for several years, Zhou Xi won the champion of artificial intelligence world competition for six times, published more than 60 articles in international top conferences and journals, and successively served as a visiting researcher in IBM TJ Watson Research Center, Microsoft Redmond headquarters Research Institute and other institutions. After that, he got the right opportunity to study in the United States.
However, in the summer of 2011, the visit of yuanjiahu, President of Chongqing Liangjiang District Green Intelligent Technology Research Institute of Chinese Academy of Sciences, prompted Zhou Xi to change the original AI exploration track. Yuanjiahu sincerely invited Zhou Xi to the Chinese Academy of Sciences in the motherland for development. Zhou Xi then returned home and founded the first AI team of the Chinese Academy of Sciences. With this foundation, Zhou Xi founded Yuncong technology company in 2015.
▲ group photo of Zhouxi (first from the left), tutor huangxutao (second from the left) and yuanjiahu (second from the right), President of Chongqing Institute of Chinese Academy of Sciences
This AI startup from the laboratory of the Chinese Academy of Sciences was first called the "national team" for this reason.
In recent years, data security has become a more and more serious problem in the field of artificial intelligence. In this regard, Zhou Xi previously replied: "because we are the 'national team', the Chinese Academy of Sciences is our shareholder, and we have joint laboratories with the Ministry of public security, the Civil Aviation Administration and the four major banks. However, we do not disclose data and attach great importance to data protection. At the same time, we have participated in the design of many national standards, which is the advantage of the 'national team'."
In this IPO, we saw that Yuncong technology company has no foreign shareholders, which is an important characteristic that distinguishes it from other leading AI enterprises.
Previously, Yuncong has obtained 11 rounds of financing, and in 2017, it received 2billion government funds from the Guangzhou municipal government; In the b+ round of 1billion financing in october2018, the investors investing in cloud include China Guoxin, Guangzhou industrial investment fund, Guangdong Yueke financial group, Shanghai Liansheng capital, Bohai Industrial Investment Fund and other Chinese national strategic investors.
According to the prospectus, before this offering, Yuncong had four state-owned shareholders, namely Nansha financial holding, Guoxin capital, Shanghai Liansheng and Guangdong venture capital.
▲ state owned shareholders of Yuncong Technology
However, the shareholding ratio of state-owned shareholders is limited. The top five shareholders of Yuncong technology are Changzhou Yuncong, Jiadu technology, Yunyi Zhongmou, dingshengxinhe and investment tycoon liuyiqian, holding 23.32%, 7.82%, 7.55%, 4.02% and 3.86% respectively. Zhou Xi is the actual controller of Yuncong technology, with 64.6% voting rights.
▲ equity structure of Yuncong Technology
Even with the support of the national team, as a highly anticipated player in the AI field, the IPO path of cloud technology is not smooth.
Wearing the title of "Ai four little dragons", Yuncong is one of the first computer vision companies to launch the listing of the sci-tech innovation board. It has started listing counseling since August 2020. However, the IPO of Yuncong technology took more than 700 days. It was suspended in March, 2021, and was included in the list of "China's military industrial complex enterprises" by the U.S. Treasury Department in December of the same year. It can be described as continuous twists and turns.
Now, the boots finally landed. In this listing, Yuncong issued 112.43 million shares at a price of 15.37 yuan, with a total share capital of 740.6706 million shares. This youngest AI four little dragons enterprise reaped a share price increase of more than 50% as soon as it opened.
Yuncong technology finally realized its IPO seven years after its establishment, moving towards a new stage of more mature development.
The annual revenue exceeded 1billion yuan, accounting for 80% of the smart governance business. Shenzhen has 36 major projects in two years
Listing is not only a glory, but also means a bottle of "makeup remover" for AI enterprises. The business performance is put on the table for market review.
AI technology R & D investment is large, and the effect is slow. The profitability of AI unicorns is worried by the market. In this regard, Zhou Xi said in a recent interview: "in combination with the current business plan and the situation of orders on hand, the company expects to turn losses into profits in 2025." This is the first time among the four AI dragons that a player has predicted the turning point of profit and loss.
From the perspective of operation, cloud technology has rapidly grown into a "1billion yuan" camp player.
As shown in the figure below, the revenue of the cloud in 2017, 2018, 2019, 2020 and 2021 was RMB 65million, RMB 484million, RMB 807million, RMB 755million and RMB 1076million respectively.
▲ revenue, net profit, R & D expenditure and gross profit margin of Yuncong technology in 2017-2021
Meanwhile, the R & D expenditures of Yuncong technology in 2017, 2018, 2019, 2020 and 2021 were RMB 59million, RMB 148million, RMB 454million, RMB 578million and RMB 534million respectively; The net profit of Yuncong in the corresponding period was -59million yuan, -148million yuan, -454million yuan, -578million yuan and -534million yuan, which was still a loss but narrowed.
Finding a field suitable for AI technology landing and carrying out saturation attack is the top priority of AI enterprises at present, and it is also a topic of concern in the industry.
In which fields does Yuncong enclose land?
Starting from computer vision algorithm, cloud technology was first anchored in the financial field at the beginning of entrepreneurship.
"The first business is to help Haitong Securities develop an identity card authentication system for remote users." Zhou Xi recalled in a recent interview. At present, in this field, the cloud has covered more than 400 financial institutions, including six major state-owned banks.
As the financial industry has higher requirements for scene complexity and security credibility than many other industries, cloud has been quickly applied across borders to smart governance, smart travel, smart commerce and other fields from the successful landing experience of technology in the financial field.
As shown in the figure below, at present, smart governance has always been the main business segment of cloud Cong, with revenue reaching 469million yuan, 434million yuan and 865million yuan respectively, increasing from 60.08% to 80.78% of the total revenue.
▲ 2019-2021 cloud revenue distribution in various application fields of Technology
From 2019 to 2021, cloud technology won a number of large projects. The contract objects are mainly smart governance (smart city) plates such as smart IOT platform, video big data system and portrait recognition software system. It can be clearly seen that government related projects are dominant. As of December 31, 2021, the company has signed artificial intelligence solution projects with a contract amount of more than 30million yuan, human-computer cooperative operation system projects with a contract amount of more than 5million yuan, or other important projects as follows:
▲ important sales orders of Yuncong Technology
For example, Yuncong is cooperating with the administrative examination and approval Bureau of Sichuan Tianfu new area to build a 476million yuan digital city 2020-2021 action plan project, which covers Tianfu brain, smart IOT platform, smart portal, digital city and other contents. It is expected to be completed before this year.
According to the prospectus, the top five customers of Yuncong technology in 2021 are Sichuan Tianfu new area administrative examination and approval Bureau, Guangzhou Nansha District Health Bureau, Beijing Huizhi Lingyun data technology company, Hengyang Municipal Public Security Bureau and Jinkai company. In 2021, the top five customers accounted for 69.58% of cloud's current operating revenue, while in 2020, the proportion was 27.92%, which shows that cloud's dependence on key customers is becoming stronger. On the one hand, the identity of the national team makes it easier to win government projects. On the other hand, the cloud has become more dependent on government key customers.
▲ top five customers of cloud Cong Technology
It can be seen that Yuncong technology is more "integrated" in sales projects, that is, it acts as a sub supplier of an integrator in the field of intelligent governance, which is similar to the model of security companies such as Hikvision and Dahua.
In fact, in addition to the large projects disclosed in the above IPO documents, cloud Cong also frequently appears in many small and micro projects in the form of integration.
According to our statistics, the 6.83 million yuan full target view analysis practical application platform project of Anyang Public Security Bureau in December, 2021, the 770000 yuan face cross network comparison platform project of Shanghai Railway Public Security Bureau in October, 2021, and the 17.06 million yuan computer equipment and software project of Zhungeer banner Public Security Bureau in September, 2021... These projects all use cloud Cong related equipment, covering video analysis server, face recognition algorithm module Snapshot library server and other products.
In general, these smart governance projects focus on security needs.
Let's take a look at the production mode. Cloud technology mainly adopts the JDM (joint design manufacturer) mode.
JDM mode means that the company puts forward product requirements and plans. Based on the basic hardware products of outsourcing manufacturers, intelligent development parts such as AI algorithm and application development of the company's leading products, as well as core links such as ID design of some products, outsourcing manufacturers produce products according to the company's needs and orders. In addition, OEM (original equipment manufacturer) mode and ODM (original design manufacturer) mode are also important production modes of cloud technology. Under the OEM mode, the company independently develops and designs relevant equipment, which is manufactured by outsourcing manufacturers; ODM mode is OEM production by outsourcing manufacturers according to company orders.
▲ main production modes of Yuncong Technology
According to the prospectus, the top five suppliers of Yuncong technology in 2021 are zhongshixun Communication Construction Co., Ltd., Shenzhen Guodian Technology Communication Co., Ltd., addition and subtraction information technology (Shenzhen) Co., Ltd., and the procurement content mainly includes servers, camera equipment, public cloud, construction services, etc.
▲ top five suppliers of cloud technology
It can be seen that unlike Kuangshi and other AI peers who adopt the mode of software and hardware self-development, cloud Cong technology emphasizes more self-development in algorithms and software.
▲ cloud slave human-computer cooperative operating system and industry intelligent solution
The company raised 1.728 billion yuan to invest in the human-computer collaboration system of cloud heavy industries and lay out the future
The sci-tech innovation board is a window for small and medium-sized enterprises with insufficient profits but scientific and technological innovation ability in China.
With the landing of the sci-tech innovation board, the cloud will take off the mysterious cloak from the AI technology concept of science and technology, as well as the core technology and innovation behind it, and will be understood and examined by more people.
Yaozhiqiang is the co-founder of Yuncong technology. In an interview in November 2021, he talked about the development stage of the AI industry. He believes that the industry will go through three stages, just like three waves:
The first wave: the breakthrough of single point technology represented by computer vision and speech recognition gave birth to the preliminary application of AI in specific scenes;
The second wave: AI has experienced a focus on single point technology. Customers gradually find it difficult to respond quickly to their complex needs, and turn to seeking comprehensive AI solutions to enable AI in the whole business chain and form an industry value closed loop;
The third wave: with the deep integration of AI and physical industries, AI will take the revolutionary improvement of user experience as the main driving factor. AI will try to embed all business processes with human and machine intelligent interaction, connect online and offline data, reallocate intelligent traffic, and greatly optimize the collaborative experience between human and intelligence.
"At this stage, the AI industry has entered the end of the first wave and started the era of the second wave of technology. Based on this, cloud technology has created the core of AI Ecology - Man-machine cooperative operating system." Said yaozhiqiang.
▲ cloud from man-machine cooperation operating system and application product development process
What is a man-machine cooperative operating system?
According to the IPO documents, the cloud slave human-computer cooperative operating system is a product that combines a computing algorithm platform with a knowledge platform.
Many people know about the deep learning framework, such as Google tensorflow, Facebook pytorch, Baidu PaddlePaddle, etc., which are tools to help people quickly build deep learning models. At present, the cloud has never independently developed a deep learning framework, but has built a man-machine collaborative operating system based on the deep learning framework.
The cloud Cong human-computer cooperative operating system was officially unveiled at the world Internet Conference in november2020. Specifically, it is not only a computing resource scheduling engine, algorithm warehouse and supporting data platform, but also a knowledge center to help customers complete complex decisions. According to the description of Yuncong's IPO document, the functional type of Yuncong's human-computer cooperative operating system is equivalent to Huawei's modelarts+modelarts pro+mindspire, Baidu's AI knowledge center, or the open-minded brain++.
However, the cloud technology man-machine cooperative operating system has three characteristics:
(1) The operating system has ecological aggregation effect: it is the core hub among hardware, software and applications.
(2) The operating system has industry characteristics: different industry platforms can be deployed for different industries, and different industry knowledge and business flows can be embedded.
(3) The operating system has the ability of cross-border integration: the single point technology cannot realize the expert knowledge structure, which requires the closed-loop of the algorithm scheme industry of the human-computer cooperative operating system. It needs to replicate the core competence of top experts, realize the ability iteration from ordinary practitioners to senior experts, and hand over the structured work to machines.
▲ main architecture of cloud technology human-computer cooperative operating system
The human-computer cooperative operating system and application products provided by the company to customers include 17 core components and applications, such as intelligent cloud platform, integrated biometric system, cloud eye face recognition service software, facego face recognition engine, fire eye face big data platform, and commercial insight platform.
According to the IPO documents, the main business of cloud technology can be divided into human-computer cooperative operating system and artificial intelligence solutions. At the present stage, artificial intelligence solutions still account for 90% of the business of cloud technology. The annual revenue of the man-machine cooperative operation system from 2019 to 2021 will reach 183million yuan, 237million yuan and 136million yuan respectively, accounting for a relatively small proportion.
▲ distribution of main business income of cloud technology from 2019 to 2021
Although the direct revenue accounts for a relatively small proportion, with the development of artificial intelligence technology and the deep integration with the real industry, the cloud human-computer cooperation operating system will play a more important strategic role, which can be described as the layout of the future.
It can be seen from the IPO documents that the primary task of this cloud fund-raising is to realize the human-computer cooperative operating system upgrade project, with an investment of 813million yuan. At the same time, Yuncong will also invest 831million yuan in the light boat system ecological construction project, which is a general service platform built based on the cloud from the man-machine collaborative operating system. In terms of functional applications and application modes, similar products include Kuangshi Pangu aiot view integrated application platform, Huawei cloud application platform Roma, etc.
In general, the investment in the man-machine cooperative operation system and the light boat platform exceeded 1.6 billion yuan, exceeding 1.4 billion yuan in the artificial intelligence solution integrated service ecological project.
▲ main purposes of the cloud from this fundraising
With the gradual disappearance of China's demographic dividend, the integration of artificial intelligence technology into the real industry has become a general trend.
According to the data of CCID consultants, a well-known research institution, it is estimated that by 2022, the scale of China's artificial intelligence industry will reach 262.15 billion yuan, with a growth rate of 28.8%. McKinsey, a research organization, predicts that by 2030, artificial intelligence will contribute an additional $13 trillion to the global GDP.
▲ scale and forecast of China's artificial intelligence industry from 2017 to 2022 (CCID consultant)
In the artificial intelligence track in the next decade, if the cloud continues to occupy the key positions, the man-machine cooperative operation system must be the key position for it to settle down.
Behind this, Yuncong technology and its subsidiaries have 316 related patents. Among them, there are 133 invention patents, 55 utility patents and 128 design patents. Yuncong has formed the R & D structure of "three joint laboratories and five R & D centers". As of December 31, 2021, Yuncong technology has 1120 employees, a decrease of 38% compared with 1799 a year ago.
▲ R & D architecture of cloud Cong Technology
Zhou Xi once said that Yuncong respects talents very much, and the company's internal working environment has been interconnected. The office in which employees work is completely based on everyone's preferences. Some people like the first tier cities and think that the quality of life there is high. Others like the second tier cities and think that the house price is relatively cheap. The company hopes to create a comfortable working atmosphere for everyone.
▲ employee composition of cloud Cong Technology
Conclusion: the AI four dragons have landed in the IPO industry to usher in a new stage
After more than 700 days, the AI national team player cloud finally landed on the scientific innovation board from technology. The successful listing means that cloud technology has been recognized in terms of scientific and technological innovation ability and business fundamentals, but it also means that it will face new challenges.
Can Yuncong really turn losses into profits in three years, as Zhou Xi said? Can the man-machine cooperation system of Yuncong support it to remain invincible in the track competition in the next decade? These questions require time to answer. We expect that this youngest AI four little dragons enterprise will usher in a new stage of higher quality development after seven years.