BEIJING, Dec. 19, 2025 /PRNewswire/ — Bairong Inc. (the “Company”, “we” , “us” or “our” ; HKEX: 6608), a leading cloud-based AI turnkey service provider, today unveiled its enterprise AI Agent strategy where it introduced its Results as a Service (RaaS) business model and launched the Results Cloud platform, an enterprise-level AI agent tailored for diverse roles and applications. Concurrently, Bairong announced it will collaborate with industry partners to build an AI Agent Productivity Ecosystem, designed to accelerate the large-scale deployment of AI agents in high-value roles such as marketing, customer service, HR, and legal affairs.
AI is not merely a tool, it should deliver tangible results in the era of AI-human co-governance.
At the launch event, Shaofeng Zhang, Founder and CEO of Bairong, stated: “The next frontier of enterprise AI isn’t about talking more or delivering more features—it’s about doing more and delivering results. RaaS empowers enterprises to transform AI into real productivity through role-specific and end-to-end execution, measurable delivery, and audit-traceable accountability. We are opening this foundational AI framework to ecosystem partners to jointly advance AI agent management—enabling the production, training, deployment, and evaluation of agents while ensuring results are verifiable, reusable, and scalable.”
AI-human co-governance as a measurable new governance model
The ultimate purpose of generative AI is to unleash human creativity. Bairong proposes a corporate philosophy of “AI-human Co-governance where Intelligence Empowers Success,” establishing a new business model where AI Agents and humans collaborate through complementary strengths. This isn’t just a slogan—it’s a measurable system focusing on the AI-agent to human ratio, workflow reengineering rates, and the maturity of AI to collaborate. In this model, humans focus on strategy, creativity, and emotional connections, while standardized, process-driven, and computationally intensive tasks are handled by efficient and tireless AI agents. Through its RaaS strategy and Results Cloud platform, Bairong empowers every enterprise to easily build and manage several AI agents, transforming AI’s potential into growth momentum, competitive advantages, and innovation sources. AI isn’t just a tool—it’s accountable for business outcomes.
Strategy and Brand Launch: BR = Best Results
The evolution of Bairong is a journey of relentless pursuit centered on delivering quantifiable business outcomes to clients. Zhang notes, “Our journey began with a simple conviction that while everyone sells hammers, clients seek gold—business results.” Over the past decade, the software industry has experimented with several models. The models range from initial project-based delivery with high customization costs and limited scalability to the sale of tools, features, and SaaS subscriptions, yet many failed to resolve the persistent issues of eroded value proposition, product homogenization, and cutthroat competition.
In the current era of the AI agent, capabilities are exploding, yet true productivity remains trapped in outdated frameworks. Bairong is driving this paradigm shift by not just building AI tools – its improving productivity. AI must deliver tangible results, not follow the old path.
“Results-driven and results-based valuation” forms the core philosophy guiding Bairong’s development. Distinguishing itself from most AI companies that merely sell features and tools, Bairong emphasizes offering value to clients through tangible business outcomes. Accordingly, Bairong has positioned its corporate strategy as “delivering AI agent solutions for results-driven clients,” revamping its brand identity to champion the “Best Results” philosophy.
Results Cloud Platform: Three-tier Architecture
The Results Cloud platform has been officially launched and serves as the core of the RaaS strategy, enabling enterprises to deploy and manage AI agents at scale, similar to hiring and managing human staff. The platform features a three-tier architecture:
Layer 1: “Baiji (百基)” — AI Infra, comprising computing infrastructure, inference engines, and domain-specific AI models.
Layer 2: “CybotStar (百工)” (a Chinese term for “hundred crafts”) — Agent OS, an enterprise-grade intelligent operating system;
Layer 3: “Baihui (百汇)” — Agent Store, featuring Bairong’s flagship AI agent workforce.
The Results Cloud platform has established an unprecedented enterprise-level AI productivity system. Its three-tier architecture not only provides a technical foundation but also addresses accountability for results through five core capabilities, elevating AI from a tool to a productivity driver.
- Observability and Measurability: The platform establishes a comprehensive monitoring system for AI agent operations. At the “CybotStar (百工)” layer, enterprise-level capabilities are separated into components, enabling end-to-end observability and fine-tuning to ensure instant compliance upon deployment while the “Baihui (百汇)” layer provides a value-driven closed-loop mechanism to track operational processes and quantify business outcomes. This allows clients to clearly track ROI per investment. This deep traceability transforms AI agent performance from a black box into a transparent dashboard, delivering on the accountability for results commitment.
- One-Click Optimized Evaluation: The platform pioneers the “One-Click Evaluation Optimization Release” experience. Leveraging the agent DevOps framework, enterprises can rapidly breakdown complex business capabilities into separate components, with the system automatically handling model selection, parameter tuning, and performance validation. The platform innovatively incorporates “Reflective Learning” technology, which transforms historical interaction records generated during agent operation into abstract knowledge through reflective processing. This knowledge is then securely processed in batches into the agent’s knowledge base, enabling automated improvements in its performance. This mechanism ensures the agent becomes increasingly accurate with use over time, eliminating the need for manual debugging and truly delivering the efficient “One-Click Optimization, One-Click Release” experience.
- Autonomous Self-Optimization: The platform’s Agent Runtime capability enables agents to achieve autonomous self-optimization in real-time. Through “Reflective Learning” technology, agents automatically distill historical interaction experiences during operation, transforming them into abstract knowledge and injecting it into their knowledge base to continuously improve performance. Compared to traditional methods that rely on manual prompt engineering, this autonomous self-optimization ensures a more stable and scalable performance enhancement for agents. As agents continuously learn and evolve with use over time, their performance can approach that of supervised fine-tuning (SFT) models requiring extensive labeled data, yet at significantly reduced costs. This truly realizes “evolution through operation” in agent lifecycle management.
- Built-in Pricing and Revenue Sharing: The platform has innovatively implemented a built-in pricing and revenue sharing mechanism at the “Baihui (百汇)” layer, transforming accountability for results into an actionable business model. The platform supports three value exchange modes: task-based pricing (for individual or batched tasks), position-based compensation (equivalent to a monthly/annual salary for a silicon-based employee), and value creation-based sharing (sharing profits with clients proportionally to performance improvements). This mechanism enables clients to manage their AI agent workforce as flexibly as human resources, creating a value closed loop where “usage is measurable, outcomes quantifiable, and returns tightly linked to value.” It truly converts AI’s potential into quantifiable business growth.
- Reducing Life-cycle Management From 2 Months to 2 Weeks: The platform has revolutionized agent development and deployment cycles through its Agent DevOps framework, reducing the traditional months-long process to just weeks. This breakthrough stems from the platform’s comprehensive lifecycle management: from unified abstraction at the LLM Ops model layer, to separated processes via Agent Builder, and continuous evolution during runtime. The entire process achieves automation and standardization. Each agent meets performance benchmarks upon deployment, eliminating additional debugging. This transformation from one-time project delivery to sustainable and ongoing productivity assets which significantly enhances the efficiency and success rate of enterprise AI applications.
These five core competencies of Results Cloud collectively form a solid foundation for enterprise-level AI deployment, transforming AI agents into manageable, quantifiable, and priceable productivity assets. This evolution elevates AI from a tool to an AI agent productivity driver that yields tangible results, ushering in a new era of ‘AI-human co-governance.’
“The Results Cloud platform integrates full-stack GenAI capabilities, built upon our ongoing services to 8,000 institutional clients and a vast number of 2C users. It supports a results-oriented, outcome-as-a-service business model for AI agents,” said Michael Wang, Chief Product and Marketing Officer and Vice President of Bairong.
Technical Base: Why Bairong Can Scale Results Delivery
Bairong’s ability to achieve large-scale delivery of results stems from its profound expertise and innovation in AI technology. The BR-Proactive LLM delivers twice the ROI of general large models in real-world scenarios, while the BR-Voice end-to-end speech model achieves a fourfold speed improvement. The BR-Vision-Doc visual language model excels at high-level tasks including parsing, extraction, comparison, and review. The BR Vortex inference engine reduces P99 latency by an order of magnitude through multi-level caching, and optimizes chip utilization by 30% via diversified computing power.
Bairong has adopted a hybrid architecture using “cloud-based training with on-premises inference”, deploying model training on Alibaba Cloud to leverage its elastic resources for peak computing demands. Meanwhile, model inference is executed on self-developed clusters, ensuring complete autonomy with predictable performance, cost efficiency, and latency control for core business processes. The platform enables unified access and intelligent scheduling for all models, automatically matching heterogeneous computing resources while supporting dynamic model slicing and aggregation. This ensures each inference task operates on the most suitable and cost-effective resource combination.
Improving Enterprise AI agent to human Ratios: From One Platform to Two Types of AI Agents
For decades, most companies have operated under a “carbon-centric” model: software systems and tools handle process fragments, while humans stitch them together. With AI agents now taking the lead, businesses are entering a new phase of AI-human synergy —AI agent workers will shoulder greater workloads in high-touch, streamlined, and measurable roles, driving up the AI agent to human ratio across organizations.
Drawing on McKinsey’s framework for the “human-Agent-robot” division of labor, corporate roles will evolve from “human-centric” to “human-machine collaboration,” and ultimately to “Agent-driven.” Core processes and tasks that are rule-intensive, involve frequent interactions, and are measurable are best suited for AI agents to handle. In practical terms, the first two key areas for large-scale implementation are EX (internal efficiency) and CX (external revenue).
Two types of AI agents: EX + CX
In the CX (Customer Experience) domain, Bairong cultivates AI agents for external revenue generation, covering scenarios like smart marketing, customer service, and retention. These agents excel at precise targeting, personalized recommendations, and complex issue resolution.
In the EX (Employee Experience) domain, Bairong cultivates in-house tech-savvy agents to excel in specialized service roles like finance, taxation, legal affairs, and HR (including resume screening and initial interviews), serving as a super assistant.
Meanwhile, Baihui (百汇）Agent Store (Ecosystem Marketplace), serving as the gateway to Results Cloud’s ecosystem, has evolved from a’ product store ‘to an’ open platform.’ This transformation attracts third-party developers and industry ISVs to co-create and deploy AI agent agents for vertical scenarios, fostering a sustainable Agent ecosystem.
Four Flagship AI Agents: Proving Results with “Model Positions”
Baiying’s (百盈 CX) “Integrated Sales & Service” AI Agent Specialists: Ensuring Every Customer Interaction is Heard and Addresses! By building a stable top-tier team, solidifying user reputation, and converting customers into business opportunities, Bairong directly addresses the three core paradoxes in traditional sales and service operations—team attrition, reputation volatility, and opportunity loss. The annual turnover rate has dropped from over 70% to 0%, peak customer satisfaction has risen from a 16% decline to a 55% increase, and the consultation conversion rate has surged by 217%, propelling sales and service operations onto an AI agent new trajectory.
BaiCai (百才EX Smart Recruitment): AI agent Recruitment Specialist – Driving Recruitment into the Autonomous Era! Through the synergy of three product matrices (Intelligence, Smart Recruitment, and Smart Management), it achieves a closed-loop recruitment lifecycle of “precision sourcing → automated flow → intelligent decision-making,” directly addressing the three major challenges in recruitment: slow hiring, inaccurate assessments, and difficult diagnoses. The recruitment cycle is shortened from 28 days to 2 days, the resume matching rate increases from 60% to 90%, and the number of HR positions to be recruited per role expands from 5 to 20, truly ushering in the “autonomous era” of recruitment.
Baijian (百鉴EX Professional Services): Elevating Large-Scale Professional Services Standards for Global expansion! Through the Baijian professional service platform, clients gain one-stop access to cross-border establishment, compliance design, and structural planning services. The platform aggregates over 80 countries/regions and 1,000+ professionals, collaborating with AI agents under a “9:1 AI agent to human Synergy” model—where 90% of high-frequency tasks (data aggregation, verification, key point extraction, traceability, and version management) are handled by AI agents. Experts focus on critical judgment and review, ultimately reducing feedback cycles from 90 days to 14 days and project costs from RMB3-6 million to under RMB1 million, making global legal, financial, and tax services more efficient and accessible.
BaiZhi (百智EX Knowledge Production): The AI agent Companion for Professionals – Listen, Memorize, and Write with Expertise! Through integrated three-terminal collaboration (recording pen + APP + PC), it enables comprehensive data intelligence aggregation and empowers knowledge production with best practices. Addressing critical bottlenecks in the knowledge chain – fragmented information acquisition, lack of methodology in knowledge production, and prolonged delivery cycles (days to months) – it creates a closed-loop from information acquisition to results delivery. The deep report delivery cycle has been reduced from 20 days to 4 days, with 80% efficiency improvement, empowering professionals to produce professional creative outputs efficiently.
Ecology and Standard: From “Publishing Products” to “Defining Rules”
A complete AI-human co-governance ecosystem requires collaborative efforts between Bairong and industry partners. At the launch event, Bairong signed strategic agreements with multiple key partners to jointly establish the four pillars of the RaaS ecosystem:
- Standard Development: Jointly produced with authoritative institutions including the China Academy of Information and Communications Technology, the “White Paper on Enterprise-Level AI Agents (Agent) Technology and Applications”（《企业级AI智能体（Agent）技术及应用白皮书》） aims to define technical frameworks, evaluation systems, and implementation standards that will lead the industry toward standardized development.
- Basic Research: Jointly established with institutions including the Gaoling School of Artificial Intelligence at Renmin University of China, the ‘AI for Business Joint Laboratory’ focuses on cutting-edge topics such as agent cognitive reasoning and complex task decomposition, maintaining technological leadership.
- Industrial Synergies: Through deep collaborations with leading cloud providers, telecoms service providers, domestic chip manufacturers, and model partners like Tongyi Qianwen, Bairong has established a comprehensive industrial chain spanning computing power, model development, and application implementation, ensuring seamless technological integration with industry needs. The Company has partnered with Gangtise Research to explore financial scenario voice models and co-develop an ecosystem for capital markets research institutions, leveraging BR-Voice technology. Additionally, Bairong strategically invested in Side Information, co-developing the next-generation AI Native card-based financial agent “Cai Chacha财查查” for securities 2B2C scenarios using its proprietary BR-LLM. By leveraging Side Information’s extensive securities industry expertise, Bairong extends its AI capabilities and ecosystem into the securities sector.
- Industry Ecosystem: Bairong was named as the co-chair of the Intelligent Agent Innovation and Application Working Group, leading the development of the Remote Banking Intelligent Agent Standard.
Bairong’s Results Cloud platform has evolved from product release to rule definition, establishing a complete RaaS value chain that spans technical standards, industrial ecosystems, and commercial mechanisms.
Platform × Ecosystem: The Multiplier Effect of Constructing a New Business Ecosystem in the Era of AI Agents
December 18, 2025 marks not just a product launch, but the dawn of a new era. In the era of AI agents, Bairong cordially invites industry partners, clients, and investors to witness and join this grand journey of AI-human co-governance to forge a shared future. Bairong will relentlessly advance its RaaS strategy, empowering every enterprise to effortlessly build and manage its Ai agent workforce through the Results Cloud platform and broader ecosystem, transforming AI’s potential into tangible growth drivers, competitive advantages, and innovation engines.
About Bairong
Bairong Inc. is a leading enterprise-level AI application service provider, delivering AI Agent to institutional clients through its Results-as-a-service (RaaS) business model. The Company applies large language models (LLMs), reinforcement learning (RL), retrieval-augmented generation (RAG), multimodal intelligence, natural language processing (NLP), deep machine learning, digital human 3D engines, privacy computing, cloud computing and other technologies. Leveraging its proprietary BR-LLM large language model and its agent building system Bairong CybotStar（百融百工）, the Company enables clients to automate end-to-end internal operations management and external business transactions. Both “Bairong CybotStar（百融百工）” and BR-LLM have completed national regulatory registration. The Company holds over 95% of intellectual property rights among its product portfolio, and a total of 461 software copyrights and patents as of June 30, 2025. Bairong serves more than 8,000 institutional customers across industries including finance, internet, retail, communications, education and healthcare, with AI Agent services applied in scenarios such as marketing recommendation and sales collaboration, intelligent customer service and outbound calls, contract, invoice and system analysis, credit assessment and fraud prevention, claims settlement and records management, as well as recruitment and operations and training empowerment.
For more information, please visit: http://ir.brgroup.com
Safe Harbor Statement
This press release contains statements that may constitute “forward-looking” statements. These forward-looking statements can be identified by terminologies such as “will,” “expects,” “anticipates,” “aims,” “future,” “intends,” “plans,” “believes,” “estimates,” “likely to,” and the negative of these words and other similar expressions or statements. Bairong may also make written or oral forward-looking statements in its periodic reports to the HKEx, in its annual and interim reports to shareholders, in press releases and other written materials, and in oral statements made by its officers, directors, or employees to third parties. Statements that are not historical facts, including statements about Bairong’s beliefs, plans, and expectations, are forward-looking statements. Forward-looking statements involve inherent risks and uncertainties. A number of factors could cause actual results to differ materially from those contained in any forward-looking statements, including but not limited to the following: Bairong’s strategies, future business development, and financial condition and results of operations; Bairong’s limited operating history; risks associated with the financial service industry, Bairong’s ability to develop and deliver services of high quality and appeal to clients; Bairong’s ability to generate positive cash flow and profits; Bairong’s ability to compete successfully; Bairong’s ability to build its brand and withstand negative publicity; and changes in client demand and government incentives, subsidies, or other favorable government policies. Further information regarding these and other risks is included in Bairong’s filings with the HKEX. All information provided in this press release is as of the date of this press release, and Bairong does not undertake any obligation to update any forward-looking statements, except as required under applicable laws.
