Defining the era of "flexible silicon-based employment": BRYANNET's exploration and leadership in the RaaS model
As the Spring Festival holiday in the Year of the Horse approaches, the nine - day consecutive break is called the "longest Spring Festival Golden Week in history" by netizens. At this time, people are eager to return home, but still have to deal with the cumbersome adjustment of work processes. Human resource allocation is one of the difficult problems. Traditional industries are facing a huge "human resource vacuum" at this time. The hardware facilities and software services purchased by enterprises are also in a state of idleness and idle running, which poses challenges to ensuring the flexible operation of enterprises.
The latest solution has actually emerged.
The "Flexible Silicon - based Reinforcement" of Silicon - Carbon Co - governance
During the Spring Festival last year, Zhang Shaofeng, the chairman and CEO of Rong360, an enterprise - level intelligent agent platform company, found that he was a bit overwhelmed. In a company with thousands of employees, implementing flat management is a huge challenge to the energy of senior executives.
Later, in an interview, he mentioned his coping strategy: "I built my own'silicon - based' assistant through the Rong360 intelligent agent platform by dragging and dropping." It can handle some daily work according to the actual situation. For example, it can set a more scientific schedule according to the priority of work and make real - time adjustments, and handle trivial tasks such as data in statistical work and file name modification. These things can be completed just by "sending a text or voice to the silicon - based assistant", allowing managers to focus more on core affairs.
What is a silicon - based assistant? According to the definition in the "White Paper on Enterprise - level AI Agent Technology and Application" by the China Academy of Information and Communications Technology, they are "agents that can perceive the environment and take actions using tools to achieve specific goals". Simply put, they are "AI agents".
Rong360 is a pioneer in this field. On December 18, 2025, Rong360 officially launched the enterprise - level AI agent strategy, proposed the RaaS (Result as a Service) business model, and launched the Results Cloud platform and the enterprise - level intelligent agent product system.
Rong360's actual actions were actually earlier than the strategic release. Now, the application of intelligent agents like "silicon - based assistants" is no longer limited to enterprise senior executives, but is also widely used among grass - roots employees in the company. "It can be between companies, and employees can also build their own silicon - based assistants for themselves."
These AI agents are not simply a collection of codes, but standardized productivity assets precipitated in the Results Cloud. Relying on its underlying Agent OS (Baigong) and the engineering delivery system, application parties can, like scheduling flexible cloud computing power, achieve the batch deployment and precise replacement of "silicon - based reinforcements" in a very short period according to the needs of business peaks, fundamentally realizing a paradigm shift from the "rigid constraint of human resources" to the "flexible guarantee of productivity".
"It is essentially a set of machine human resource system." An overseas industry blogger named "AI Thinking Chain" summarized on X. This highly professional tweet has received wide attention, with more than 400,000 views.
Once activated, this set of machine human resource system can theoretically achieve the on - demand allocation of silicon - based employees. Zhang Shaofeng revealed that the silicon - carbon ratio within Rong360 has reached 1:150, which means that one carbon - based employee can dispatch about 150 units of silicon - based employees.
They are widely used in scenarios that can be covered by AI, such as interviews, accounting, taxation, research, and marketing. As the training ability of AI improves, the self - coordination ability among silicon - based employees is strengthened, and the number of silicon - based employees that carbon - based employees can dispatch also rises rapidly. "Silicon - carbon co - governance" is growing within Rong360 and among its customers with an irresistible trend.
The Dusk of SaaS and the Dawn of RaaS
Silicon - carbon co - governance is not only a solution, and it does not only address the single problem of the "human resource vacuum", but also means the implementation and application of a new production paradigm. At the same time, it also brings the issue of the demise of the old paradigm to the fore.
The SaaS (Software as a Service) model is the target of the challenge. The important difference in business logic between SaaS and RaaS is that the former charges a rigid service fee according to the actual number of employees of the customer, while the latter cannot charge according to the number of silicon - based employees. It charges according to the business results after the application of silicon - based employees. When the marginal benefit brought to customers by the increasing silicon - carbon ratio is stronger than that of the increasingly rigid SaaS model, "charging by headcount" becomes the last straw that breaks the back of SaaS—if I have already purchased ERP software, why do I still have to face the trouble of the "human resource vacuum"?
In the past year, more and more industry insiders have gradually become clearer about this.
On January 20, 2026, a16z, a top venture capital firm in Silicon Valley, released a in - depth analysis report on AI applications, which put forward a view that has been hotly discussed in the industry: the logic of the SaaS industry is undergoing a fundamental change. In the past, enterprises purchased software by headcount to assist employees in their work. In the era of rapid AI evolution, intelligent agents will directly replace the labor force to complete tasks. A large amount of data precipitated by carbon - based employees in the past is used to train intelligent agents in different scenarios—for example, using sales data and scripts to train intelligent agents can ensure that they reach the average sales level and significantly improve the conversion rate.
This discovery has actually been noticed and applied by first - class enterprises. At the CES conference in January 2026, McKinsey, a world - renowned management consulting firm, revealed that in the past 18 months, the number of its AI agents has soared from a few thousand to 25,000. Nearly 30% of McKinsey's employees are composed of "digital employees"—or "silicon - based employees". McKinsey will assign at least one silicon - based employee to each carbon - based employee in the next 18 months.
In the view of Zhang Shaofeng and his colleagues, silicon - based employees can provide measurable job performance just like carbon - based employees. Behind the silicon - based employees is the infrastructure based on the cloud platform and the corresponding services. The platform itself is deeply involved in the customer's business and takes on greater responsibilities and risks for the customer's performance. The deep coupling of its products with the customer's performance has led to a fundamental change in the relationship between the platform and the customer.
In the AI 1.0 era, SaaS was considered a common model for enterprise - level AI applications. However, as the potential for business growth brought to enterprises has been exhausted and homogeneous competition has intensified, the application of SaaS is approaching its end, and a new model is emerging. From a16z to McKinsey, they have all discovered that the large - scale application of AI agents has promoted the emergence of the RaaS model.
In Zhang Shaofeng's words, RaaS is not selling shovels to customers for them to dig gold by themselves, but using the same shovel to help customers dig out the gold. Rong360's Results Cloud was developed and created according to this thinking. Its intention is obvious: to align with business results in the RaaS model, transplant the business experience and capabilities precipitated by customers over the years to AI agents, and thus elevate the "assistance" of silicon - based employees to carbon - based employees to "silicon - carbon co - governance".
Rong360's confidence in its technical capabilities supports its RaaS strategic ambition. Results Cloud has a three - tiered structure to meet different levels of needs, namely AI Infra ("Baoji", computing power infrastructure + inference engine + exclusive domain AI model), Agent OS ("Baigong", enterprise - level intelligent agent operating system), and Agent Store ("Bahui", which includes Rong360's flagship silicon - based employees). It is worth noting that in addition to acting as a "product store", Agent Store can also be upgraded to an open platform to expand the silicon - based employee ecosystem with third - party developers.
After breakthroughs in the R & D of four major technical fields, namely conversational AI engineering, voice understanding and synthesis, retrieval enhancement and long - term memory, and tool call and orchestration, the development of Rong360's silicon - based employees has now achieved engineering, and the deployment cycle has been shortened from two months to two weeks.
This means that whether an enterprise has sudden needs or is seeking incremental growth in regular business, it can quickly expand its labor force through Results Cloud to meet the enterprise's different needs. Thus, Rong360 has become a real "AI labor dispatch company".
But it seems to go beyond "labor dispatch". While providing silicon - based employees, Rong360 has introduced its unique "Silicon - based Employee Home" management system to carry out standardized management of these employees. Each silicon - based employee has a unique employee number, performance indicators, and is equipped with a carbon - based "workplace mentor" to ensure that they meet the standards once they start working. In this management system, the application of silicon - based employees has greatly simplified the training process of carbon - based personnel and the system running - in process, and the flexible way of silicon - based labor dispatch also saves enterprises a regular IT service subscription cost.
On this basis, Results Cloud applies silicon - based employees in different scenarios and summarizes the "Four Flagship Silicon - based Positions", namely Baiying (integrated sales and service), Baicai (intelligent recruitment), Baijian (professional service), and Baizhi (knowledge production). Baiying is oriented to the customer experience field, covering scenarios such as intelligent marketing, customer service, and retention, bringing business growth; the others are oriented to the employee experience field, penetrating into professional service scenarios such as finance, taxation, law, human resources, and research to improve efficiency.
Judging from the statistical data, the silicon - based specialists in the four positions deserve the title of "flagship". The consulting conversion rate of Baiying has increased by 217%, Baicai has shortened the recruitment cycle from 28 days to 2 days, Baijian has reduced the legal, business, finance, and taxation costs totaling millions of yuan to less than one million yuan, and Baizhi has shortened the delivery cycle of in - depth reports by 80%. The four flagship positions work internally and externally to initially form a closed - loop for enterprise operation, bringing silicon - carbon co - governance to a higher level.
Defining a New Benchmark for the AI Era with a New Paradigm
The RaaS model has initiated silicon - carbon co - governance, changed the way of enterprise governance, and brought about a win - win situation.
The platform is undoubtedly the direct beneficiary of silicon - carbon co - governance. RaaS is centered on "result - orientation and result - based pricing", including pricing by position, pricing by output, and profit - sharing by performance, which is more flexible than the previous SaaS subscription payment model. As customers have achieved obvious growth, in the first half of 2025, the revenue of Rong360 also increased by 22%, and the gross profit margin remained stable at 73%.
Importantly, in the RaaS model, customer data comes from private domain traffic, and its AI training is different from the general AI large - scale models of large companies. AI agents trained through private domain traffic can continuously improve the fault - tolerance rate over time, strengthening the coupling relationship between the intelligent agent platform and the customer's business, and building a solid moat for the continuous construction of the RaaS model.
This difference has also led to a shift in the cognitive coordinates of the industry and the capital market. Their attention has shifted from traditional SaaS evaluation factors such as software effort and customer unit price to indicators representing new - quality productivity such as the "number of active silicon - based employees in a productive state" and "work share" to judge the binding degree between the intelligent agent platform and customers and the revenue growth prospects.
The year 2025 is considered the "Year of AI Agents". The domestic industry is seeking to formulate evaluation specifications for the application effectiveness of enterprise - level AI agents. Other standards regarding the operation and compliance of AI agents are also being introduced one after another. Some listed companies choose to disclose their AI agent application situations. In several major AI enterprise mergers and acquisitions in 2025, a new evaluation system oriented to the RaaS model is being established.
If the data intuitively shows the benefits brought by RaaS to customers, then Rong360's attempt is bringing implicit and conceptual changes to the industry. More than a hundred years ago, Ford's assembly - line revolution greatly improved labor productivity, but it also made people fall into the instrumental rationality of only pursuing value and ignoring people's emotional and spiritual values. The evolution of AI technology has given rise to RaaS, shifting this work rationality to "loyal, efficient, and tireless" AI agents. This is not about silicon - based employees replacing carbon - based employees, but about liberating carbon - based employees from repetitive and high - frequency work, giving them more time and space to engage in creative work.
In this sense, Rong360 is promoting a revolution in a new production paradigm. It is not only an AI agent platform company, but also a pioneer and rule - maker in the global RaaS field.