Chinese research service debuts AI agent for future industry analysis

The interface of the "Tianyan" intelligent research agent is pictured. [Photo provided to China.org.cn]

China Economic Information Service (CEIS) on Thursday unveiled "Tianyan," an intelligent research agent designed to help analysts study emerging industries during China's 15th Five-Year Plan period (2026-2030).

The system, a major achievement of CEIS' "AI + think tank" initiative, is designed to understand, reason and adapt over time, providing end-to-end intelligent support for industrial decision-making.

Zhang Chunfang, product manager at the Digital Economy Research Center of CEIS, said the platform is built around three core capabilities.

First, the platform can automatically map enterprises across industrial chains. Using a proprietary algorithm to identify industrial correlations, the system links and dynamically tracks upstream and downstream enterprises. It currently covers 17,873 key enterprises in six strategic sectors: quantum technology, biomanufacturing, hydrogen and fusion energy, brain-computer interfaces, embodied intelligence and sixth-generation mobile communication.

Second, Tianyan features a proprietary knowledge engine for future industries. The development team continuously feeds it with first-hand materials from frontline research and field studies, ensuring every assessment produced by the system is evidence-based and traceable.

Third, the system incorporates an autonomous industrial assessment engine. Rather than searching data manually, it draws on modular expert capabilities — industrial-chain analysis, policy interpretation and enterprise profiling — to automate complex industrial evaluations.

Cao Jie, deputy general manager of the Economic Think Tank at CEIS, said Tianyan will initially serve two main use cases.

For government decision-making, the system can generate comprehensive industrial profiles, regional comparisons and capability assessments to inform policymaking. For investors, it can flag promising early-stage, small-scale and hard-tech enterprises through a dedicated evaluation model.

Cao Wenzhong, president of CEIS, said the organization is accelerating the deployment of intelligent agents in high-value scenarios, including industrial assessment, policy analysis, risk alerts and digital government operations.

He added that CEIS plans to build a collaborative, self-improving network of such agents to bring think tank services further up the value chain, support national strategies and industrial innovation, and establish a flagship national economic information platform.