Stories about QCon
4 related stories
失败一次，受益多次：测试 Agent 从记忆到持续学习的工程实践｜QCon上海
AI InsightThe iteration and maintenance of test scripts is a long-standing pain point in automated testing, requiring new solutions. Zhao Xuefeng's team is exploring the use of AI to autonomously explore interfaces, structure fuzzy requirements, and generate scripts and execution plans, such as using knowledge bases, scripts, and execution products to design Agent workflows and implement failure-driven upgrade self-healing loops.Key TakeawayTest script iteration and maintenance requires new solutions, Zhao Xuefeng's team is exploring the use of AI to autonomously explore interfaces.Why It MattersThe iteration and maintenance of test scripts is a long-standing pain point in automated testing, requiring new solutions. Zhao Xuefeng's work may have a significant impact on the field of test automation, such as improving test efficiency and reducing costs.Who's Affected- BeneficiaryTest AutomationThe iteration and maintenance of test scripts is a long-standing pain point in automated testing, requiring new solutions.
What's NextIn the future, test automation may rely more heavily on AI and machine learning technology, Zhao Xuefeng's work may become a reference case for the industry.Importance 70/100从 Harness 到 Loop：Agent 如何走向可靠交付与持续进化？
AI InsightThe core competition of AI Agents is shifting from foundational models to peripheral engineering systems. The transition from Harness to Loop represents a shift from static architectures for single-task execution to dynamic closed-loop systems capable of self-evolution driven by execution feedback, fundamentally reshaping human-agent collaboration.Key TakeawayThe competitive boundary of Agents is shifting from model capability rivalry to the construction of self-evolving closed loops in engineering systems.Why It MattersSingle metrics like code generation rates fail to solve the transfer of engineering costs. Only through Loop architectures enabling continuous feedback and self-evolution in production can the reliability bottlenecks of Agent adoption be truly broken.Who's Affected- BeneficiaryAI R&d TeamsIntroducing Loop engineering architectures and automated Eval systems can significantly improve the reliability and delivery efficiency of Agents in production.
- BeneficiarySoftware Engineering Infrastructure VendorsDemand for infrastructure supporting Harness and Loop, such as trace-level observability, tool protocols, and security sandboxes, will surge.
What's NextSubsequent observation should focus on practical enterprise cases of deploying Loop architectures in production, especially the maturity of automated Eval, security sandboxes, and self-evolution mechanisms based on execution feedback.Importance 80/100从 Agent 到 Loop：AI 正如何重构软件工程？｜QCon 上海 2026 即将开幕
AI InsightTechnical practices from leading Chinese internet companies in 2026 show that AI applications are shifting from single-point model invocations like coding to comprehensive Agent ecosystems covering R&D, infrastructure, and complex consumer chains. This implies that AI's core value is evolving from tool-based efficiency to complete service fulfillment and business restructuring.Key TakeawaySoftware engineering is shifting from isolated model calls to end-to-end restructuring centered on Agents and continuous loops.Why It MattersAs AI penetrates complex consumer chains such as travel, the challenge shifts from model reasoning to connecting dynamic supply and service networks, determining whether AI can truly enter transaction and fulfillment stages.Who's Affected- At RiskSoftware Engineering TeamsR&D organizations must adapt to process changes from writing code to building continuous Agent loops.
- BeneficiaryConsumer Internet PlatformsIntroducing AI-native service networks may restructure connections between humans and services, improving fulfillment efficiency.
What's NextSubsequent focus should be placed on technical architectures and data flywheel cases in consumer-level Agent deployments presented at QCon Shanghai 2026 to verify commercial viability.Importance 78/100Agent 的记忆不在对话里：把企业数仓沉淀为可治理的共享语义记忆｜QCon上海
AI InsightThe construction of AI Agent memory is evolving from personalized conversational history to shared semantic governance of enterprise data warehouses. This shift indicates that the core barrier for vertical multi-agent systems is no longer remembering user preferences, but how to transform fragmented enterprise knowledge into structured assets with verifiable writing and conflict resolution.Key TakeawayAgent memory focus is shifting from conversation history to shared semantic governance of enterprise data warehouses.Why It MattersCurrent mainstream memory frameworks focus on conversational interactions, failing to meet strict data accuracy requirements in industrial scenarios. Transforming enterprise data warehouses into governable shared semantic memory directly determines whether multi-agents can scale from experiments to production environments.Who's Affected- BeneficiaryEnterprise AI DevelopersProvides engineering paths to transform enterprise data warehouses into structured shared semantic memory, helping reduce deployment error rates in vertical domains.
- WatchingAgent Framework VendorsEncourages existing conversational memory frameworks to expand capabilities toward enterprise data semantic governance and integration.
What's NextSubsequent observation should focus on whether architectures based on data warehouse semantic governance in enterprise agent deployments can effectively resolve hallucinations and error injection in complex business queries.Importance 65/100