As AI-assisted development becomes mainstream, traditional code organization and development workflows are increasingly insufficient. Large-scale codebases, multi-repository environments, and legacy enterprise systems introduce significant challenges for applying AI effectively.

This session introduces Spec-Driven Development (SDD) as a practical approach to scaling AI-assisted engineering. Instead of relying on ad-hoc prompts, SDD treats specifications as first-class artifacts that drive code generation, validation, and evolution.

The talk focuses on real-world challenges and solutions in large-scale environments:

• Codebase Structuring: designing repository and module structures that are AI-friendly and spec-aligned
• Cross-Repository Coordination: managing dependencies and changes across multiple repositories using spec-driven workflows
• Enterprise Integration: integrating SDD into existing systems, CI/CD pipelines, and development processes

Speakers:

Zhangjian He: Huawei Cloud Senior Engineer | Huawei Cloud Open Source Team Member

Open-source enthusiast and experienced engineer. Since starting software development in 2017, I have played a key role in leading the cloud-native transformation of Huawei Cloud’s IoT platform and the development of an industrial IoT platform. I serve as an Apache BookKeeper/ServiceComb PMC member, Apache Pulsar Committer, and CNCF openGemini Maintainer. I specialize in coding, software engineering, open source, and ecosystem development, driving best practices and development standards.