Recently, GF Securities has keenly captured the new trend of the securities industry's transformation from "tool empowerment" to "intelligent reconstruction," taking the lead in completing a comprehensive AI-native upgrade of its "Easy Gold App," achieving a leap from "comprehensive trading service tool" to "round-the-clock intelligent investment partner," and launching a new paradigm of intelligent services featuring "personalized experiences and instant gratification."

**Technology Foundation: Self-developed "Tianji Zhirong" Large Model Directly Addresses Financial AI Implementation Pain Points**

For a long time, brokerage apps have commonly suffered from issues such as bloated size, complex information, complicated interactions, and insufficient personalization. As a result, investors often face challenges in information screening and high barriers to professional tools during the investment process.

With the rapid development of large model technology, the demand for AI applications in the securities industry has become increasingly urgent. However, general-purpose large models have limited adaptability in the financial sector, prone to issues such as insufficient professionalism, missing business logic, and compliance risks.

Addressing this problem, a relevant official from GF Securities pointed out: "To truly solve industry pain points, AI must become the business hub, not a functional add-on."

Based on this philosophy, GF Securities launched localized large model application practices from early 2025, constructing the "Tianji Zhirong" model service as the technical foundation for the native upgrade of the "Easy Gold App."

The "Tianji Zhirong" large model combines GF Securities' years of practical data and professional investment research capabilities, integrating RAG and Agent orchestration technologies to form a controllable, reliable, and optimizable localized intelligent hub. This model not only possesses financial knowledge reserves and logical reasoning capabilities but can also efficiently implement in various business areas such as market data, trading, information, and accounts within the compliance framework, achieving comprehensive AI reconstruction from underlying architecture to interaction methods. This solves the past functional redundancy and interaction barriers of brokerage apps, lowers the threshold for investors to access professional services, and avoids issues such as insufficient professionalism and compliance risks that general-purpose large models face in financial scenarios.

**Three Core Transformations: Reconstructing Intelligent Service Logic to Create Next-Generation AI-Native Securities App**

The next-generation AI-native "Easy Gold App" has undergone core design and reconstruction using the "Tianji Zhirong" large model and AI technology from the underlying architecture of securities trading and investment services to user experience.

Beyond technical upgrades, the revolutionary aspect of its service logic lies in "making AI understand intent rather than users adapting to functions," treating AI and large model services as the core engine driving the entire Easy Gold investment service system, completely reshaping the interaction between users and financial data, market analysis, and trading decisions.

First, the next-generation "Easy Gold App" builds a proactive personalized engine through full-scenario AI empowerment, deeply embedding AI capabilities in high-frequency service scenarios such as market data, trading, information, and accounts throughout the entire process. It has created over ten AI intelligent agent modules around the complete investment journey, including i-Assistant Easy Xiaotao Plus, i-Monitoring, i-Trading, i-Review, i-Information, i-Advisor, i-Smart Investment, i-Account, etc., providing users with precise services adapted to different scenarios and forming a closed-loop ecosystem covering "perception-analysis-decision-execution-review."

Moreover, it achieves breakthrough in interaction modes through "conversational revolution," directly reaching professional answers through AI thinking chains, allowing investors to complete information acquisition and investment operations simply by "talking and chatting."

The launched large model version investment assistant "i-Assistant Easy Xiaotao Plus" differs from traditional apps' passive responses. Relying on five core capabilities including thinking chain reasoning, metacognitive reflection, multi-layer analysis, intelligent knowledge invocation, and reasoning process visualization, it can provide friend-like proactive "conversational services" 24/7 in over 20 core scenarios such as stock selection and stock diagnosis, significantly improving investors' information acquisition efficiency.

Furthermore, GF Securities has achieved the virtuous cycle of "AI serving as a super computing engine while human experts focus on decision-making to empower customers" through its localized intelligent hub - the "Tianji Zhirong" large model service.

This comprehensive upgrade of the "Easy Gold App" is not only a product innovation but also sets a new benchmark for brokerage services, redefining the value connotation of intelligent investment services, evolving investment services from "cold tools" to digital partners that "understand finance, understand markets, and understand you better."

The "Easy Gold App" has evolved from "changing sails on an old ship" to an "intelligent vessel," ushering in a new era of "intelligent investment partner ecosystem."

In the future, GF Securities will continue to use the "Tianji Zhirong" large model service as its intelligent hub, leading the transformation of intelligent service models in the securities industry, continuously strengthening technological foundations, deepening the application of AI technology in financial services, enhancing scenario empowerment, co-building industry ecosystems, and helping the securities industry's digital development reach new heights.