A person who invested in medicine a few years ago came to invest in AI, and in his eyes, AI is the same as innovative drugs, burning hundreds of billions of dollars without even burning a slag.
In his opinion, CPO is similar to photovoltaic inverters, both of which are used to smash foreigners' jobs at low domestic prices.
A person who invested in cloud computing before came to invest in AI, and his cognition, AI is no different from SaaS, API is the primary productive force, and there is as much intelligence as there is labor.
A second-level dog with assets of less than 6 million also got over the wall GPT, he thinks, he is not much different from the "rich family" who can't sell the old house, AI is another ordinary person's outlet after Taobao and Douyin, grab it and go ashore, after all, there are 40 places left in one lesson a week, and if you don't grab it, it will be gone.
A veteran who has been doing finance for 10 years has also done self-media, and he found that the people who sell AI courses make more money than Ultraman himself, and Leeks doesn't care what they talk about, so I'm going to start a live broadcast. These financial veterans, a 20,000-word AI report at their fingertips, and a roadshow of 400 people are no problem, but when they turn on a live broadcast, they can't understand what they say, and they don't even seem to have a normal IQ when they make video clips like PPT.
The question of the conversion of medicine to AI: whether there is any profit after burning money, the question of new energy to AI: whether there is any that can make money from domestic customers, the question of cloud computing to AI: whether there is any head of the code, the question of the second-level dog to AI: whether there is a company that can take the group purchase channel for me, and the financial veteran ask: Is there any such AI company that is not covered by our group's previous report?
There really is, but AI companies like ¥ Bairongyun-W (6608.HK) that meet the conditions are often more exciting in the Futu comment area than the seller's report: it was all Zhuge before the low suck, and Meng Shu became Meng after the high sell-all is a real hero. Moreover, the anchor who sells AI courses does not provide stock recommendation services, and the lower right corner is hung with a small yellow car, not a small windmill.
Now there are some research titles that say Bairong Cloud: a one-stop service AI technology leader, which interprets the MaaS and BaaS businesses from the official website, but retains the impression of "only empowering the financial industry and not diversifying the layout". In fact, this understanding is biased.
If you read the following description of Bairong Cloud's MaaS business, are institutions and retail investors stupid: MaaS (Model as a Service) means that Bairong Cloud provides various AI model functions and gives customers in various industries an API interface, and they can call AI services by themselves.
MaaS first appeared in 2012, when AI had not yet entered the era of large models, but the theory of MaaS had already taken shape. It is divided into the basic layer, the middle core layer, and the bottom extension layer. Among them, the basic layer is the most basic information that covers the user, the middle core layer describes the important user characteristics, such as the user's preferences and goals, and finally the underlying extension layer contains personalized knowledge with user characteristics, such as the user preferences obtained through semantic relationship analysis, and the personalized model network. Finally, through the basic layer to the middle core layer, and then to the lowest expansion layer, a MaaS platform is formed. This platform can be plugged into the cloud as a stand-alone service platform, between the PaaS layer and the SaaS layer. Therefore, MaaS can be used to make up for the shortcomings of SaaS enterprises, mainly to make up for the shortcomings of SaaS that are difficult to customize according to the personalized needs of customers.
Experience from the first perspective how Bairong Cloud's customers use MaaS services: for example, an enterprise, (identity is a financial institution, e-commerce platform, logistics, leasing, etc.), hopes that the products or services they provide can have a better user experience, and it is best to finally realize AI automatic recommendation. In the past, in order to achieve AI personalized recommendations, they needed a large amount of data, computing resources and a professional team to train AI models and deploy them, but when people in HR and procurement departments looked at the department's budget, they chose not to be scolded by the boss between reducing costs and increasing efficiency, and AI is getting farther and farther away.
At this time, if there is an AI company, it can encapsulate its various AI models and functions into API interfaces, so that these enterprises can call them from the cloud, which not only helps them save computing power and development, but also makes a lot of money, which is why Bairong Cloud does MaaS business. With cloud AI capabilities, these enterprise customers can dream of one day wrestling with giants.
In the first step, enterprise users log in to the API interface to find AI models and services suitable for their business, and they can choose a trained and optimized "small model" from Bairong Cloud's MaaS platform, so they do not need to train AI models from scratch, saving a lot of time and money. In the second step, enterprise users will use the interfaces and tools provided by MaaS service providers to integrate personalized recommendation models into their business processes, including but not limited to bank credit systems, e-commerce platforms, logistics management systems, etc. Through a simple API API call, enterprises can obtain the recommended results of Bairong Cloud's AI model and finally display it to users.
In the process of operation, Bairong Cloud, as a MaaS service provider, is responsible for real-time inference and update of AI models, as well as Fintune. This means that SMEs no longer need to worry about the complexity of model training and inference, they can focus on their own business logic and platform optimization.
By leveraging Bairong Cloud's MaaS services, enterprise customers can quickly deploy personalized recommendation functions to display relevant products or services based on user preferences and historical behaviors. Not only does it improve the user experience, but it also increases sales conversion and customer loyalty. The number of API calls corresponds to the call fee to be paid to Bairong Cloud.
According to the 2023 annual report just released by Bairong Yunchuang, the company will achieve a total revenue of 2.68 billion yuan in 2023, a year-on-year increase of 31%, and an adjusted net profit of 375 million yuan, a year-on-year increase of 28%. Excluding the book loss due to the impairment of goodwill of the acquisition, the adjusted net profit was as high as 436 million yuan, exceeding market expectations. Among them, as a MaaS business that can support the daily peak of 300 million calls of B-end institutional customers, the revenue was 891 million yuan, a year-on-year increase of 17%, the core customer retention rate was as high as 99%, and the average income of core customers increased to 3.495 million yuan.
Bitter for those investors who have tapped out Bairongyun's AI capabilities, they are competing with a bunch of SaaS impression institutions, and the band has been done, but it may also be about to schizophrenia.