On December 5-6, 2018 Chuangye State 100 Future Leaders Summit and Chuangye State Annual Conference (hereinafter referred to as "Chuangye State 100 Future Leaders Summit") was held in Beijing National Convention Center. On the afternoon of December 5, Du Jianghua, executive vice president of Qiniu Cloud, delivered a speech on the enterprise service forum with the theme of "making cloud+AI become the standard configuration of enterprise services", sharing three views on providing cloud services for future enterprise development:
New IT architecture=cloud computing technology+big data technology+artificial intelligence technology
In the era of data explosion, a one-stop data management solution is needed to make cloud and AI complement each other
All kinds of enterprise services will gain new advantages through intelligent transformation
(Du Jianghua, Executive Vice President of Qiniuyun)
The following is a transcript of the speech:
Thank you very much for the opportunity to share with us. I want to share with you how to combine cloud and AI from the perspective of a technology service company. The previous guests mentioned that AI has become the standard configuration of enterprise services. From my perspective as a cloud service startup company, we believe that cloud+AI will become the standard configuration of enterprise services. I remember that in 2014, when we shared with others, we usually talked about making cloud services the standard configuration for entrepreneurship, and then said that our cloud computing would be as popular as highways and power grids. Four or five years later, this description has become a reality, and cloud computing has really become an infrastructure.
In that year, we talked about how to transform our traditional IT into cloud computing. The most important thing we talked about was to go to IOE, remove pure hardware services such as IBM, ORACLE, EMC, and become a cloud computing service. With the development of big data and AI in recent years, we believe that the new IT must combine cloud computing technology, big data technology and AI technology to form a new IT architecture. If you are in the industry, you may know that the new chairman of Alibaba released the latest organizational structure adjustment of Alibaba last Monday, November 26. The biggest change is that Alibaba Cloud has become Alibaba Cloud+Smart, which also verifies the direction of our technology development. We believe that in the process of shaping new IT, what kind of company can become the trendsetter of new IT architecture, everyone has the opportunity, and there are many uncertainties.
Therefore, cloud+AI will inevitably become an enterprise service, and companies in both entrepreneurial and mature fields will be used as a standard option. What is the relationship between the two? Last month, General Secretary Xi of the Shanghai International Expo mentioned the Science and Technology Innovation Board. The most important thing is that AI enterprises will have a great opportunity to give priority to support. Therefore, there is no doubt that AI will become a national strategy in essence. Large and small companies, whether mature companies, listed companies or start-ups, are moving in this direction. All kinds of enterprise services can gain new advantages through intelligent transformation. Two days ago, I talked with my friends about the intelligent HR of some companies. By observing the various data performance of employees on the intranet in the recent period, we can judge the possibility of their resignation within a month or two. This is very helpful. Intelligentization can help judge the potential trends of employees.
Thirdly, AI has a very strong industry scene attribute. Just now, two guests mentioned that there is no standard product in the industrial Internet era, so we must do a lot of customized things for industrial products and process transformation. It is bound to involve the AI mentioned in the fourth point. In fact, AI cannot exist without the cloud. It must have the ability of massive computing and data management. Only in this way can cloud+AI complement each other, promote each other, and become a good helper of enterprise services.
You know what kind of data explosion era we are in now. More and more interactive methods have produced more and more massive data, including mobile phones before, cameras now and other products, and AR/VR and other interactive methods in the future. Our future society is composed of massive data, that is, there must be new IT suppliers. With the development of the IoT Internet and the extensive visualization of interactive forms, 90% of the future network traffic will be IP based video data, which is quite consistent with the previous guests' views. So the explosion of data requires us to have a new way of data storage and analysis. If such a company can not only provide video support capabilities but also intelligent capabilities, not only provide reliable support in storage, computing, and distribution, but also can bring continuous analysis capabilities to these videos, upgrade video technology, and tap more value, It must be very meaningful. In fact, all industries need the overall solution of data management. I will take the security industry as an example. The country has invested heavily in the security industry. In fact, if you have the ability to structure data, through video analysis, you can predict the possible security risks in a certain area, which can greatly reduce the input of police personnel. In addition, data storage is a huge challenge. There are more and more cameras. It is reported that a district in Shenzhen has replaced more than 200000 cameras in a few months. In addition, for the security industry, the pan social scenario still lacks a comprehensive cloud based solution. At present, it is still a localized solution, and there is huge space here.
Qiniu Cloud is such a company. At the beginning of its establishment, the essence of the services we provided was data management around entrepreneurship projects and APP data services. In the past two years, we have undergone transformation. We have gradually become a cloud service based on the Internet to provide innovation around massive data, A one-stop solution service provider with big data AI as the core to help customers manage massive data. It emphasizes "one-stop". Let me share our current data volume. Through the past six or seven years, we have managed more than 200 petabytes of data, and the number of cloud files has reached more than 200 billion. We have also received investment from many well-known domestic investment institutions.
In fact, the most important thing is one-stop. Many companies may provide storage services, or data distribution services, AI services, and big data services. However, during the development and transformation process, Qiniu Cloud has accumulated flexible, innovative and efficient one-stop services around massive data. The essence of this one-stop service can be used as an independent module to provide services, or it can be combined arbitrarily to form the model that customers want. From content generation to data mining, data processing, data distribution, and presentation to consumers, it provides a complete solution in the entire value chain. We will focus on how we integrate with AI, what AI technology we integrate, and what advantages AI technology has. Because the advantage of Qiniu Cloud is based on massive Internet data, including all kinds of video and image storage data, we can enrich all kinds of scenes through computer vision on this basis, including yellow identification, terror detection, scene recognition OCR recognition, face recognition, object recognition and other various AI services also provide the data platform AVA platform that supports algorithms. Users can do a variety of algorithms on it for enterprise AI innovation exploration. We provide such capabilities to output. In essence, big data products focus on operational log data and various log data on the server side and the operation and maintenance side, including massive Internet log data to provide real-time data analysis. Whether online or offline, real-time analysis and retrieval can be done to help decision makers extract the data value they need.
What are the advantages of our AI? First, the algorithm is powerful. Generally, AI companies only emphasize how powerful their algorithm teams are, such as how many doctors and professors they have. However, we not only have powerful algorithms, but also have the ability to calculate, analyze and process massive data. Based on our years of experience in public cloud and our algorithm team, we have set up artificial intelligence joint laboratories with Fudan University, Huashi University and other universities. Because we have massive data accumulation and more reliable data dependence on the algorithm model, we have more advantages in computing power and algorithm. We have an in-depth learning platform, open source system for everyone to use and can give a feedback. We are also an open ecology. We will attract all kinds of developers and institutions to do more innovation based on massive data on our platform.
Let's share with you what vertical industries we focus on, because you know the industrial Internet, and we will also use our first mover advantage to accumulate data on the Internet. In addition to our most advantageous mobile Internet field, we have a large number of customers, many of whom have been listed in various capital markets. At present, we have deep customers in the radio and television media field, such as Zhejiang News Media, and many well-known customers in the intelligent hardware field, such as OPPO, fluorite, etc. They not only integrate our storage, data processing AI has made comprehensive integration. In addition, we have accumulated a lot of cases in the financial and government fields in the last two years, which will not be described in detail here.
Finally, I want to tell you a very specific case. During the Expo last month, the state leaders visited Pudong, and we made a product demonstration to the state leaders. One of the most troublesome things in Pudong when managing the urban environment is the disposal of muck trucks. We judge whether his driving route is compliant by combining the camera on the road and the license plate number. If the caravan fails to cover the muck firmly, the system will give an alarm. After the alarm, the staff can set up a card along the route he takes in advance to see if he violates the rules. After four weeks of treatment, the number of violations per month was 80 times, and then it was reduced to 25 times. Four weeks later, eight weeks later, the violation never appeared again. It became 0. This is AI's ability to upgrade the industry and help our law enforcement departments create greater social value. This is also the last thing I shared. What we have done recently is industrialization, and what we will do in the future is more urban brain, which will give greater play to our value in security.
That's all for my sharing. Thank you!
