On October 22, at the 2026 China Computer Conference, Alibaba's Jiang Guanjun, head of AI algorithms at the Qianwen division, shared the technical practices behind the creation of Personal Agent. He believes that the true stage of Personal Agent should bring about the creation of incremental value and personalized proactive services, with model, memory, and ecology as the three essential elements in building this core capability.
Jiang Guanjun divides the development of AI applications into three stages: dialogue, agency, and assistant. Currently, AI applications are rapidly advancing towards the assistant stage, where they begin to autonomously plan and execute around user goals by integrating professional experience and personal context, in addition to delivering specific tasks, and further identifying needs that users may not have noticed or overlooked.
The Development of Personal Agent Will Enhance User Experience
He stated that the theoretical upper limit of Personal Agent in solving daily work, learning, and life problems is close to 100%. The key to unlocking this potential lies in better matching system capabilities with problem-solving paths, finding optimal solutions for different issues through the efficient synergy of model, memory, and ecological capabilities.
At the model level, Qianwen places greater emphasis on the stability of task execution. To this end, Qianwen has established a user demand perspective covering over 30 leading industries, identifying more than 8,000 atomic model capability requirements, and involving industry experts in model evaluation and optimization to continuously enhance the model's ability to understand real user needs and complete complex tasks.
At the memory level, Jiang Guanjun indicated that Personal Agent needs to understand users over the long term, recognizing the connections between historical information and current needs, and avoiding the noise and reasoning costs that come from simply accumulating dialogues. For example, AI can combine historical dialogue about pension situations mentioned by elderly users to provide advice on employment issues for their children that is more aligned with the family's circumstances. Qianwen designs its memory system by analyzing event relationships and stores and retrieves information hierarchically based on the strength of these relationships, improving the accuracy of memory usage.
At the ecological level, Jiang Guanjun believes that as model capabilities and Harness technology continue to develop, the ability to connect and collaborate within the ecosystem will become key to the development of Personal Agent. Qianwen is building a model-centric open service system that integrates professional knowledge, tools, industry skills, and third-party agents, connecting services such as e-commerce and local life, as well as professional capabilities in fields like healthcare and education, to enhance the execution of complex real-world tasks.
The synergy of the three elements ultimately points to proactive services: memory provides personal context, the model identifies potential needs and plans actions by combining professional knowledge, and the ecology supports execution. For instance, in health management, it continuously monitors rehabilitation arrangements and their impact on daily life, while in financial management, it reminds users to pay attention to new risks and opportunities.
At Qianwen, the demand for complex tasks and proactive services continues to grow, with the average Token consumption for complex tasks increasing by 1.6 times over the past three months. Recent news indicates that Qianwen has launched a new feature called "My Applications" for its personal AI assistant "Qianwen Go!" Users can easily build applications for life, work, and learning scenarios simply by stating their needs, without the need for programming. This feature is currently open for invitation testing.
Item Specification Processor Not disclosed RAM Not disclosed