CompanyGraph
Range Intelligent Computing Technology Co., Ltd.
300442 · SZSE · China
Builds AI algorithms trained on legally non-exportable Chinese 5G and industrial data, embedding them into state-owned infrastructure under mandatory domestic sovereignty constraints.
China's Cybersecurity Law and Data Security Law require that AI systems processing Chinese citizen and critical infrastructure data be developed, trained, and deployed entirely within Chinese computational infrastructure, which forces Range's algorithms to be embedded directly into customers' real-time manufacturing control systems and telecom traffic management layers rather than hosted at arm's length. That embedded position triggers MLPS certification and Cyberspace Administration algorithmic approval obligations that any replacement provider must revalidate over a six-to-twelve-month cycle, locking switching costs into a regulatory requalification process that Range's own entry already completed. Because inference and deployment replicate cheaply across additional facilities and base stations once a model exists, the binding limit on growth is not compute but the acquisition of engineers who hold expertise in both AI systems architecture and China-specific regulatory compliance together — a combination concentrated in a narrow tier-one city talent pool that cannot be outsourced or automated. The same legal prohibition that prevents competitors from replicating Range's data position also prevents Range from moving it, so a tightening of algorithmic disclosure requirements or a reclassification of training datasets under stricter state-security provisions could suspend Cyberspace Administration approval across the entire domestic deployment base at once, with no jurisdictional alternative available.
How does this company make money?
Money flows in through software licensing for AI algorithm deployment across customer manufacturing facilities and telecom infrastructure, through subscriptions for cloud-based model inference services, and through professional services engagements covering custom algorithm development and regulatory compliance consulting.
What makes this company hard to replace?
AI models are embedded into customers' real-time manufacturing control systems, and replacing them requires six to twelve months of requalification cycles for any new algorithms. Integration with China Mobile and China Telecom's proprietary 5G network management protocols creates additional technical switching barriers. MLPS cybersecurity certifications must be revalidated for any new AI software provider serving state-owned enterprises.
What limits this company?
The Cyberspace Administration's Algorithmic Recommendation Management Provisions require disclosure of training data sources and model decision logic for every AI system processing Chinese citizen data, and MLPS certifications must be revalidated on each new deployment. The throughput bottleneck is therefore not compute or capital but the acquisition of engineers who hold expertise in both AI systems architecture and China-specific regulatory compliance — a combination that cannot be outsourced or automated and concentrates in a narrow tier-one city talent pool.
What does this company depend on?
The mechanism depends on access to China's domestic semiconductor fabrication capabilities for AI inference chips, compliance certifications under China's Multi-Level Protection Scheme for cybersecurity, integration partnerships with China Mobile and China Telecom for 5G network access, approval from the Ministry of Industry and Information Technology for telecommunications software deployment, and infrastructure from Baidu AI Cloud or Alibaba Cloud for model training and deployment.
Who depends on this company?
Chinese state-owned manufacturing enterprises rely on real-time AI-driven quality control systems, and their production line optimization would degrade without them. China Mobile's 5G network performance management depends on predictive algorithms for traffic routing. Chinese financial institutions using algorithmic trading systems would face execution delays if the low-latency AI inference platforms became unavailable.
How does this company scale?
AI model inference and algorithm deployment replicate cheaply across additional manufacturing facilities and telecom base stations once a model is developed. The bottleneck as the company grows is engineering talent acquisition in China's tier-one cities, because algorithm development for telecommunications and industrial applications requires expertise in both AI systems and China-specific regulatory compliance that cannot be easily outsourced or automated.
What external forces can significantly affect this company?
US-China technology export controls restrict access to advanced GPU architectures needed for large-scale AI model training. China's Common Prosperity policy drives government preference for domestic technology providers over foreign alternatives. Demographic aging in China's manufacturing workforce is accelerating demand for AI-driven industrial automation systems.
Where is this company structurally vulnerable?
The same Data Security Law prohibition that prevents replication also prevents portability. If China's regulatory regime tightens algorithmic disclosure requirements further or reclassifies the training datasets under stricter state-security provisions, the Cyberspace Administration approval on which every customer deployment depends could be suspended or revoked — collapsing access to the entire domestic deployment base at the same time, with no jurisdictional alternative available.