Hua Jiankan
Product Lead of CodeFlicker at Kuaishou.
Responsible for the product strategy, development, and deployment of the intelligent development assistant **CodeFlicker**, dedicated to enhancing engineering efficiency, experience, and developer well-being through AI-powered R&D tools. With extensive hands-on experience in the AI DevOps domain, previously worked at Ant Group and Tencent. Led the end-to-end creation (from 0 to 1) of cloud-native and AI-native R&D products, building intelligent development platforms and successfully driving large-scale adoption in enterprise environments.
Topic
The Era of Collaborative Software Development: Exploration and Practice of L3 AI Coding
AI programming tools are undergoing a paradigm evolution from L1 single-point code assistance, through L2 development assistance, to L3 collaborative development, with technical bottlenecks and cognitive challenges intertwined along the way. This talk will draw on insights from real engineer user scenarios to analyze the core challenges in moving from L2 assistance to L3 collaboration, presenting a three-pronged approach—concepts, tool design, and best practices—to explore breakthroughs and provide actionable lessons and inspiration. Outline: 1. L1–L5 Capability Layers in AI Programming: From Assistance to Collaboration, the Intelligent Evolution of Software Development 2. Exploration and Implementation of the AI Collaborative Development Paradigm 2.1 Insights from Real User Scenarios, Applied Back to Real Scenarios 2.2 Developer-Centric Principles for Designing AI-Native Development Tools 3. Enterprise-Scale Efficiency Evolution at Kuaishou: AI Programming Driving Systematic Productivity Leaps Across Tens of Thousands of Developers 4. Reflections and Insights for the Future
