As the mid-year reporting season draws to a close, the performance of the AI sector is becoming increasingly clear. Against the backdrop of high investment in generative AI technology and an unstable business model, coupled with tightened spending by downstream customers, AI companies that can maintain stable profits or even achieve growth are particularly valuable.
Bairong Yunchuang is such a rare AI company. Its financial report for the first half of the year shows that through the strong performance of its AI main business, the company not only achieved an adjusted net profit of 197 million yuan, but also achieved a net profit margin of 15%, successfully joining the ranks of a few AI technology companies that continue to make profits. What is particularly eye-catching is that Bairong Yunchuang's gross profit margin has continued to remain at a high level, reaching 73% in the first half of the year, which is particularly rare.
The main reason why Bairong Yunchuang can continue to maintain a high gross profit margin and achieve year-on-year growth in net profit is due to its mature AI business collaborative operation model that has been running smoothly: the company's two core businesses - MaaS (Model as a Service) and BaaS (Business as a Service) develop in conjunction, especially the BaaS business, which has risen as the company's second growth curve, relying on generative AI technology, and achieved a revenue of 900 million yuan during the reporting period, a year-on-year increase of 11%.
I. Why has generative AI become the core engine of profit growth?
Why has generative AI become the core of Bairong Yunchuang's profit growth? There are mainly two reasons:
First, the mature and efficient business model has brought stable growth momentum to the revenue side.
Bairong Yunchuang's BaaS business uses generative AI technology to provide one-stop services for financial institutions and non-financial institutions. This business accurately matches and filters users for stratification, and uses AI intelligent voice robots, SMS services, manual and other methods to reach users, helping commercial institutions achieve efficient new customer marketing, old customer revitalization, and intelligent operations.
In the exploration of AI business models, there are mainly two major landing scenarios: B-side and C-side. Although the threshold for C-side scenarios is low and commercialization seems easy, in reality, due to the influx of a large number of players and homogeneous competition, the commercialization results of C-side are not ideal, so Bairong Yunchuang chose to aim its AI business at the B-side, which has greater potential commercial value compared to C-side.
However, establishing a stable and sustainable profit model is extremely challenging. B-side applications usually face limited fault tolerance and high performance requirements, so they are often willing to purchase services only when they see results, which leads to a very high entry threshold for B-side applications.
So how does Bairong Yunchuang do it?In the domestic BaaS sector, most companies tend to pay for the actual revenue generated. However, BaiRong YunChuang has abandoned this billing method in favor of a strategy that charges fees only when it genuinely helps clients achieve revenue growth. Unlike other robot suppliers who charge based on call duration or other usage metrics, this model directly links service effectiveness to revenue, providing clients with a more secure and efficient service option. With the courage to guarantee performance results and significant achievements in actual operations, BaiRong YunChuang has achieved a revenue growth rate far exceeding the industry average, laying the foundation for profit growth.
Secondly, the high investment period has passed, alleviating pressure on the profit side.
For AI companies, exploring generative AI technology is a task full of opportunities and immense challenges. The initial development and application of new technology not only incur high costs but also carry significant risks—once a company decides to invest, it is like embarking on an irreversible journey, making it difficult to withdraw or change direction easily. Moreover, high investments in the short term often do not directly translate into significant profit growth, thus bringing long-term financial and fiscal pressure to the company.
Compared to many companies that hastily joined the large model competition after the rise of ChatGPT last year, BaiRong YunChuang's BaaS business has already taken a step ahead, with its development history dating back three years. Therefore, it has now passed the most fanatical and high-investment stage of the industry's early period and entered a more stable and rational development phase.
Furthermore, BaiRong YunChuang has not blindly followed the trend of pursuing high-concept, large-quantity models. Instead, based on its actual situation and market demand, it has created large models early on and, on this basis, used Mixture of Experts (MoE) technology to develop a series of lightweight models targeting specific fields. By flexibly combining expert models from different fields, BaiRong YunChuang has ensured high-performance services while effectively controlling costs and improving overall cost-effectiveness.
Now, as the high-investment period gradually passes, generative AI has become the core engine for BaiRong YunChuang's profit growth. Its efficient business model, combined with forward-looking strategic layout, has brought the company continuous and stable revenue growth and improved profitability.
II. Crossing the Key Inflection Point of the J-Curve
However, it is worth noting that if we observe the financial reports horizontally within the industry, many companies claim that "revenue brought by generative AI is increasing"; but in reality, there is a significant difference in performance among different companies. Some companies have adopted a continuous high-spending strategy, trying to quickly expand market share by "burning money." This strategy is essentially a high-risk "overdraft" behavior that may lead to irrational market prosperity. In contrast, companies like BaiRong YunChuang have steadily crossed the key inflection point of the J-curve growth model, showing a positive cycle and significant economies of scale.
Regarding the J-curve theory, we need to briefly review it: due to the advanced technology, high R&D costs, and gradual market acceptance, the profit changes in the AI industry naturally follow the typical J-curve growth trajectory. This process includes the introduction phase, growth phase, maturity phase, and a possible decline phase for specific industries or products. In the introduction phase, AI companies face the dual challenges of technological complexity and market uncertainty, requiring substantial investment in R&D, team building, and market expansion. This stage often comes with significant financial pressure because revenue growth often lags behind cost expenditure, leading to a decline in profits. However, once AI technology gains widespread market recognition and achieves scaled profitability in practical applications, the company crosses the inflection point and enters the growth phase. In this stage, its revenue and profits will grow rapidly, starting a new chapter of high-speed growth.
As early as 2022, BaiRong YunChuang crossed the break-even point of net profit, starting the trajectory of profit growth. In the first half of 2024, against the backdrop of adverse downstream demand, BaiRong YunChuang's generative AI business still showed strong resilience, offsetting the adverse effects, which is also a side proof.According to the performance report, in the first half of the year, Baishun Yunchuang's BaaS financial industry cloud revenue reached 589 million yuan, with a year-on-year growth rate as high as 20%. In the insurance industry cloud field, Baishun Yunchuang's new policy/single/renewal premium scale reached 1.9 billion yuan and 970 million yuan respectively, with a year-on-year surge of 103% and 47%. This series of data indicates that Baishun Yunchuang's generative AI revenue is rapidly expanding.
Baishun Yunchuang has not only ushered in a positive cycle of financial models and scale effects, but also built a virtuous cycle of R&D and growth with the emergence of the technology flywheel effect. As the scope of technology application continues to expand, the company can gather a more massive training dataset, which plays an important catalytic role in the deep optimization of AI models. The continuous optimization of technology and the surge in data volume complement each other, jointly driving the acceleration of the company's technology model flywheel effect.
In the first half of this year, Baishun Yunchuang has made significant progress in natural language processing (NLP) technology, upgrading traditional NLP technology to more advanced proactive large model technology. This upgrade has strengthened the intelligent voice interaction chain by integrating model quantization, distillation, and separated inference architecture technologies, providing users with more efficient and accurate service experiences.
Baishun Yunchuang has also introduced Retrieval-Augmented Generation (RAG) technology to further improve the accuracy and credibility of large models. RAG technology enables the model to understand the context more deeply and effectively integrate the retrieved information into text generation to meet actual needs. In the RGB benchmark test conducted by third-party evaluation institutions, Baishun Yunchuang's large model demonstrated excellent performance in multiple key capabilities, with an overall accuracy rate slightly exceeding ChatGPT3.5, proving the company's professional strength and leading position in the AI technology field.
The intelligent voice robot Voice GPT has also achieved significant technical breakthroughs. Based on the same source technology as ChatGPT, Voice GPT's response time has been reduced to less than 500 milliseconds, while introducing emotional recognition and emotional voice output functions, raising the accuracy of semantic understanding to more than 97%. These upgrades have enabled Voice GPT to play a key role in intelligent customer service, credit card and financial marketing, customer callbacks, and other links, and have increased its daily AI outbound call capacity to 50 million times, greatly improving service efficiency and customer satisfaction.
By abandoning the money-burning strategy and cultivating a dual positive cycle of finance and technology, Baishun Yunchuang has built a foundation for long-term stable growth and avoided the "bubble."
Looking forward to the future: the potential of the vertical scene digital intelligence service market is huge.
Looking ahead, with the rise of the millennial generation and Generation X, customers' dependence on digital tools is increasing, and investment habits, interaction patterns, and service needs have undergone fundamental changes. China's wealth management industry is undergoing unprecedented changes. Baishun Yunchuang relies on generative AI technology and provides customers with efficient vertical scene digital intelligence services by solving pain points with its unique business model and technical advantages.
In the first half of 2024, Baishun Yunchuang continuously expanded the application scope of its BaaS business model and successfully empowered customers in various fields such as banking, small and micro enterprise operations, and wealth management. Among them, in the banking field, the digital intelligence green financial solution jointly created by Baishun Yunchuang and Huaxia Bank was selected as an excellent case by the Ministry of Industry and Information Technology, fully demonstrating its leading position in the field of green finance. At the same time, in the wealth management business, Baishun Yunchuang effectively activated the potential of banks' wealth management business by providing digital intelligence customer group operation solutions for many banks, achieving rapid growth in customer asset scale.
Through a series of successful practice cases, Baishun Yunchuang's strength in the field of digital intelligence services has been fully verified. Among them, the author noticed that Baishun Yunchuang provided a digital intelligence wealth management solution for a joint-stock bank, effectively activating the long-neglected long-tail customer group in the bank. For customers with asset management scale (AUM) mostly below 10,000 yuan, Baishun Yunchuang helped the bank change its traditional management strategy for these customers by building an efficient intelligent customer group operation system with "one triangle, one chain, and one ring" as the core. Combined with the optimization of personalized marketing language and AI outbound call technology, the product's intention rate has significantly increased, from 4% to 10%.Within just a few months, the system has helped banks successfully increase the number of customers with AUM1 of 10,000 yuan or more by 26,000 households, exceeding the control group by 19,000 households. This has achieved in-depth mining and value enhancement of existing customers, creating new growth opportunities for the long-tail market of the bank's wealth management business.
The current wealth management industry is facing fierce competition as well as opportunities of the times, thus urgently needing digital intelligence services for self-upgrade and transformation. According to authoritative institutions such as the National Bureau of Statistics and Goldman Sachs, the wealth income of Chinese residents continues to grow, and the scale of the middle-income group continues to expand, providing a broad development space for the wealth management industry. It is expected that by 2025, the total scale of investable assets of Chinese residents will reach 50 trillion US dollars and maintain a double-digit growth rate. This trend has prompted the Chinese wealth management industry to accelerate its digital transformation to meet the growing digital needs of customers.
The management of BaiRong Cloud Creation is also full of confidence in the future market space of BaaS business. The management predicts that in the next few years, in the credit scenario, the market scale of online and digital is expected to grow from 17 trillion to 45 trillion, with a compound annual growth rate of 13%. Its market space is considered ten times larger than the credit scenario. In 2022, the online scale was 130 trillion, and it is expected to reach 400 trillion by 2030.
Faced with the booming trillion-scale market, can BaiRong Cloud Creation firmly grasp this opportunity with the advantage of generative AI technology? We can wait and see.
post your comment