On April 22, Dr. Xiaohua Zhou, Founder and CEO of DolphinDB, was invited to deliver a keynote lecture at Shanghai Advanced Institute of Finance (SAIF), Shanghai Jiao Tong University, titled "Quantitative Trading Infrastructure and New Paradigms in the Age of Intelligent Agents."The lecture was open to both the university community and the public, attracting numerous students and industry professionals.

Lecture Recap

At the start of the lecture, Dr. Zhou introduced the fundamental logic and development trajectory of quantitative investing, pointing out that the core essence of quantitative trading lies in capturing market mispricing opportunities through data mining and model analysis. As industry competition intensifies, the arbitrage opportunities from simple factors continue to narrow, and practitioners face increasing demands for data depth, system performance, and strategy iteration speed. The rapid development of AI technologies is providing new solutions to these core needs.

During the lecture, Dr. Zhou analyzed AI evolution across four stages: perceptual intelligence has matured, cognitive intelligence is gradually reaching its ceiling, and the current main battlefield of AI development has shifted toward collective intelligence—multiple agents collaborating with division of labor to form a closed-loop system that operates autonomously across perception, reasoning, execution, verification, and iteration.

He noted that finance, being inherently digital, involving high-stakes decisions, and characterized by never-ending market games, is the most natural breeding ground for collective intelligence. Agentization in various specialized areas has already begun to take shape:

• Banking: Credit approval has evolved from traditional manual due diligence to second-level credit granting based on over 500 feature dimensions, with dynamic real-time adjustment of credit limits.
• Securities: Algorithmic trading now accounts for approximately 70% of trading volume in U.S. equities, with end-to-end latency in top-tier quantitative systems reaching as low as 1 millisecond.
• Insurance: Risk control agents respond to extreme risk events within milliseconds, with dynamic pricing adjusting in real time to market changes.
• Asset Management: Robo-advisors integrate customer profiles and market data into a single framework, moving from templated allocation to true personalization.

This trend is also reshaping the entire financial software industry—traditional process tools are being systematically taken over by agents. Whoever defines the agent collaboration protocol will control the pricing power of the entire ecosystem. Consequently, the financial industry is imposing higher demands on underlying infrastructure: millisecond-level response, unified stream and batch processing, high concurrency stability, and full data lineage traceability.

In response to this trend, Dr. Zhou detailed DolphinDB’s full-stack quantitative solution, covering the complete pipeline from data storage and real-time computation to strategy R&D. The system integrates over 20 stream computing engines and more than 2,000 financial-specific functions. Its unified stream-batch architecture allows research and production to share the same codebase, reducing development costs by 90% and improving data analysis speed by over 100 times compared to traditional solutions. This provides the robust, low-latency foundation that agent systems demand in production financial environments.

On building AI capabilities, Dr. Zhou highlighted several applications launched by DolphinDB for the intelligent agent era:

• DolphinX: A next-generation quantitative research platform integrating AI agent technology. It manages both general and financial domain tools through a unified MCP/Skill layer, supporting automated execution of scenarios such as strategy backtesting, factor calculation, research report reproduction, and asset pricing.
• Starfish Research Report Analysis Assistant: Automatically parses research reports, extracts factor formulas, and generates executable code, compressing the traditional multi-day process of "report → factor replication → evaluation" into a closed-loop automated execution.
• Strategy Backtesting Agent: Users describe a strategy in natural language, and the system directly returns backtest results. By decoupling AI’s "intent understanding" from the "deterministic execution" of the template system, the strategy pass rate has increased from 84% to 99%.

He emphasized that AI does not replace expertise but enhances it. DolphinDB combines VectorDB and TextDB for RAG-ready data retrieval with the MCP protocol for seamless AI-tool integration — delivering an intelligent decision-making system built for interpretability, traceability, and continuous iteration.

During the Q&A session, students engaged in lively discussions with Dr. Zhou on topics including quantitative strategy development and the practical implementation of AI in finance. Drawing from hands-on experience, Dr. Zhou addressed each question thoughtfully and encouraged students to start building their own quantitative research systems while still in school. He noted that DolphinDB's extensive tutorials, active community, and comprehensive documentation offer strong tool support and a clear learning path.

DolphinDB Cerulean Campus Program

To advance university collaboration, DolphinDB has officially launched the Cerulean Campus Program — an initiative dedicated to bridging industry and academia through collaborative innovation and joint talent development. The program aims to integrate DolphinDB into university curricula and cultivate a new generation of fintech professionals equipped with global perspectives, innovative thinking, and hands-on expertise. Collaboration formats include guest lectures, curriculum development, talent training, and joint research, offering students access to rich learning resources, internship opportunities, and research topics.

To date, DolphinDB has established partnerships with a number of leading universities across China, including Shanghai Jiao Tong University, Peking University, Zhejiang University, Fudan University, Nanjing University, Renmin University of China, University of Science and Technology of China, The Chinese University of Hong Kong (Shenzhen), and more.

About DolphinDB

DolphinDB is a high-performance distributed time-series database. Beyond efficient storage and querying of massive datasets, DolphinDB pioneers a fully featured programming language for complex analytics, alongside a high-throughput, low-latency, and developer-friendly stream processing framework — making it one of the most computationally powerful database systems available. DolphinDB is widely adopted by leading financial institutions across securities, fund management, banking, and insurance, as well as top enterprises in IoT-driven industries including energy, power, and industrial manufacturing — significantly enhancing large-scale data analysis efficiency and substantially reducing development costs.