Jintao Li

Build trading automation, risk-control, and LLM-assisted agent workflows for asset management and interbank trading systems. - Designed automated trading algorithms for an interbank trading system, translating trading rules, execution logic, and market constraints into system workflows. - Developed LLM-powered trading assistant capabilities to support trading workflow guidance, information retrieval, and task automation. - Built trading-agent style workflows that connect market context, internal rules, tool usage, and human review instead of treating LLMs as standalone chat interfaces. - Developed automated risk-control modules for investment trading systems, including rule checks, exception handling, and operational safeguards. - Used AI-native coding workflows to accelerate prototyping, system design, debugging, and iteration while keeping reliability, review, and production constraints in focus.

Engaged in research work related to network operating system, responsible for the research and development of switch web application bridge

In the context of cloud platform overselling and hybrid deployment, I am involved in online offline hybrid deployment task optimization. The work mainly includes designing resource portraits, designing tied accounting method and enhancing container isolation. After the optimization, the hybrid deployment performance was improved by 60%.

Participated in the development of internationalized e-commerce and internal platform, and completed the code migration of some sub-modules, covering 90% of the original project work with migrated code functions.

Participated in the machine learning model acceleration project, mainly responsible for the graph-related algorithm module and benchmark development, with an average acceleration performance of 20% compared to the original model.