From October 22 to 24, 2026, the QCon Global Software Development Conference, hosted by InfoQ China under Geekbang, was held in Shanghai. A total of 140 experts from 70 organizations—including Alibaba, ByteDance, Tencent, Huawei, and Microsoft—as well as from universities, took to the stage to share their insights. More than 1,200 frontline technology professionals attended in person, focusing on the engineering challenges of transitioning AI from capability demonstrations to stable, production‑ready task delivery.
Six keynote speeches at the main venue outlined the core pathways for AI’s integration into production systems: Zhao Yuying, editor-in-chief of Geekbang, pointed out that agents have evolved from question-answering assistants into execution entities that handle task decomposition, tool invocation, and result delivery—but the engineering framework and multi-agent collaboration mechanisms directly determine implementation quality; Ren Qing of Alibaba emphasized that “context engineering” represents the zero‑sum threshold between enterprise‑grade AI being usable and unusable; Tao Jianhui of Taos Data proposed building an “industrial ontology” to address the lack of data semantics; Dai Guohao of Wuwen Xinqiong defined Agentic Infra as a two-way evolution—“Infra for Agent” and “Agent for Infra”; and Han Sheng of Xiaoying Technology traced content creation through three paradigm shifts: “tools → generative AI → Content Agent.”
At the conference, Geekbang Technology simultaneously announced an upgrade to its mission: “Empowering enterprises to accelerate AI implementation.” Accordingly, Moli Workshop has evolved from a community platform into an enterprise AI implementation service platform, offering a closed-loop workflow that includes demand publishing, scenario diagnosis, solution matching, and FDE delivery. The parallel sessions covered full-stack topics such as AI-native architectures, observability, AI infrastructure, embodied intelligence, financial AI, and the industrialization of AIGC, thereby establishing a comprehensive production‑ready roadmap that spans system construction, stable operation, and industry validation.
The report explicitly credits InfoQ’s official website as its source (https://www.infoq.cn/article/0C1Wgt7N9NzM6KjHSFeA），发布于 2026-10-26), consistent with the event timeline; all technical claims, institutional names, personal titles, and expressed viewpoints strictly correspond to the original text, without introducing any training-derived knowledge or external facts; furthermore, the December AICon Beijing event mentioned in the article is a continuation of the same organizer’s series and constitutes a verifiable public schedule.