Min Zou
Algorithm Director at Kunlun Tech Co., Ltd.
Zou Min, who holds a graduate degree from the Institute of Mathematics, Chinese Academy of Sciences, is currently the Algorithm Director of the 2050 Team at Kunlun Tech Co., Ltd.. He has been deeply involved in the development of multiple domestic and international agent products at Kunlun Wanwei. His long-term research focuses on information retrieval, large models, and agent development. He is also the author of the book *AI Search: Fundamentals and Frontiers*.
Topic
How to Build High-Quality and Robust AI Agents: Insights from My Experience Developing Over 10+ Agents
Here’s a polished English version of your abstract and outline: --- The speaker will provide an in-depth look at building high-quality AI agent systems, sharing insights across multiple dimensions, including agent architecture design, context management, toolset design, and model selection. **Outline:** 1. Introduction to AI Agents (5 mins) 2. Let the Problem Define the Agent Architecture: Single Agent vs. Multi-Agent vs. Hierarchical Agents 3. Effective Context Management: Long Context vs. Context Engineering 4. Tools Matter: How Well-Designed Toolsets Improve Robustness and Execution Efficiency 5. Model Selection: Using a “Golden Dataset” to Choose the Best Model for Your Business Problem
