"Mobile cloud is undoubtedly a very important service, including mobile R&D, HTTPDNS, mobile hotfix, mobile push, mobile security, mobile testing, SMS service, mobile data analysis, mobile user feedback, etc. Hang Yu continued.

Enterprise-level mobile application R&D service for short, referred to as EMAS, is oriented to the enterprise service market, and expects to open up Jiangyan's DevOps R&D support capabilities and mobile APP basic middleware capabilities accumulated in the mobile Internet industry for many years to customers, and help traditional enterprises quickly complete the transformation and upgrading of business mobility.

EMAS includes DevOps workflow management for the full life cycle of mobile apps, opens a large number of atomic-granular middleware and infrastructure services for the five functional domains of R&D, testing, release, operation and maintenance, and operations, and precipitates solutions for five major scenarios, including continuous delivery, componentization, cross-platform development, pan-quality management, and unified network access.

Mobile testing is a service that mainly includes compatibility testing, H5 testing, ANR, crash, monkey testing, cpu time, non-CDN resources, browser compatibility testing, and other categories.

There are many versions of the Android system, coupled with the customized modifications of various manufacturers, which makes the problem of Android fragmentation very serious.

Developers need to spend a lot of money on all kinds of real devices, spend a lot of time testing on different Android devices, and the crash problem is still serious.

Compatibility testing is all about solving this problem. Users only need to upload the application package, then directly select the real device, and submit it to the cloud for testing.

The compatibility test will help users automatically conduct installation tests, startup tests, stress tests, and uninstall tests on the selected real terminals, and finally present the overall compatibility report and all the details of all crash hazards to the developer, so that the developer can experience the ultimate test service without leaving home.

H5 applications have reached a level of parity with Native applications, and more and more developers are willing to devote their energy to more convenient and lightweight H5 applications.

At the same time, issues such as compatibility and performance of H5 applications are becoming more and more important.

H5 testing provides developers with compatibility and performance tests of H5 applications on real computers, while intelligently traversing all subordinate pages of H5 applications and giving points for H5 applications to be optimized.

There are a few other testing services, so I won't cover them all.

It took more than two hours for Hang Yu to outline the functions he wanted, while Yue Qin and the others had memorized more than thirty pages, and their wrists ached.

There are more than 20 major functions and nearly 200 classification functions.

According to the chairman's plan, Weibo Cloud will become a business no less than online shopping in the future. Jiang Rong sighed.

"Yes, I hope that Weibo Cloud will become the world's No. 1 cloud service provider in the future, and as the first company in the world to complete big data and cloud computing, we have the confidence, determination, and ability to complete this plan. Hang Yu said.

This sentence refreshed Yue Qin and others' understanding of Weibo Cloud, and the attention suddenly increased by more than one grade.

Weibo cloud was also very important, but it definitely did not reach the level of online shopping business, and even Jiangyan Company's main business could not be evaluated. After Hangyu was redefined, it was instantly upgraded to the main business, almost on par with online shopping.

"To complete so much technology development, we definitely don't have enough people at our disposal, at least more than half of the technology company at the headquarters needs to be involved. Guan Yonglin said.

"It can be appropriately expanded and supplemented, but it cannot affect the technology development and maintenance of other products. It is certain that there is not enough manpower, and there is still a shortage of talents in this area in China, so you can only work hard, and train while developing and developing. Hang Yu said.

"We have been nurturing talents in this area, and our cooperation with universities has begun to bear fruit. However, it is not easy to train a qualified talent, and I propose to expand the scope of cooperation and invest in key universities in the province and even the whole country. In addition, it is better to improve the treatment of talents in this area. Guan Yonglin said.

"Yes, you can do it. Hang Yu said.

——

Guan Yonglin returned to the technology company with his notebook and meeting minutes, convened a meeting of technicians in charge of big data and cloud computing, and found that Hang Yu's request was really not ordinarily large.

Some of the service requirements put forward by Hang Yu have already been made, and some are being done. But there are still many service projects that they have not thought of, or have thought of, but do not have the human and financial resources to implement.

For example, artificial intelligence, a project that technology companies are focusing on research and development, and the only application research and development at present is intelligent customer service. However, it is much more difficult to complete intelligent applications such as Weibo cloud than intelligent customer service.

Hang Yu proposed artificial intelligence projects such as machine learning, speech recognition and synthesis, human-computer dialogue, printed text recognition, face recognition, and image search, each of which requires a deep understanding of artificial intelligence.

For example, intelligent voice interaction, which is based on speech recognition, speech synthesis, natural language understanding and other technologies, gives enterprises a "listening, speaking, and understanding you" intelligent human-computer interaction experience for enterprises in a variety of practical application scenarios.

These technologies are suitable for a variety of application scenarios, including intelligent question answering, intelligent quality inspection, real-time recording of court hearings, real-time speech subtitles, interview recording and transcription, etc., and have application prospects in many fields such as finance, insurance, justice, and e-commerce.

Needless to say, face recognition, this technology has been proposed very early, has a wide range of applications, has a good prospect, and is very sought after by the industry.

Face recognition technology can be applied to industries such as finance, security, e-commerce, smartphones, and entertainment pictures.

Specifically, it can be applied to face authentication in a series of face application scenarios, such as face payment, face card scanning, face login, VIP face recognition, face sign-in, face customs clearance, etc.

"With so many R&D projects added all at once, we can't afford to share it. Wu Zhixiao scratched his head, their work was already very busy, and when they thought of developing hundreds of projects, they suddenly had the urge to hit the wall and die. "We can't do so many projects alone, and we have to strengthen cooperation with universities across the country," he said. ”

"The most important thing is the money, we are not only short of people but also lack of equipment, and we have to expand immediately. It also costs money to cooperate with universities and other research institutes, otherwise why would people help us with research. Chai Jin said.

"Collaboration is important, but you can't always expect others, because universities and other research institutes are not very efficient. Instead of giving them money, we should expand our own strength. Guan Yonglin said.

"Prioritize first, or we'll have no clue. Wu Zhixiao said.

Because there were so many projects added, they didn't know where to start for a while.