The center of gravity in AI has shifted — from standalone models to autonomous agents that reason, act, and collaborate. And at nearly every layer of this new stack, open source is doing the heavy lifting: open table formats and data infrastructure feeding the models, open agent frameworks orchestrating the reasoning, and open protocols letting agents talk to tools and to each other.

Drawing on real projects across the Apache ecosystem and beyond — from Apache Iceberg for AI-ready data, to open-source agent SDKs like Strands, to emerging standards such as MCP — this keynote traces how a healthy open source community becomes the foundation for production-grade agentic AI. We’ll look at what breaks when agents move from demo to production (state, control, observability), how the community is solving it in the open, and where developers can plug in and contribute.

You’ll leave with a clear mental map of the modern AI + open source stack, and concrete ways to build on it — and give back to it.

Speakers:

Betty Zheng: Senior Developer

Yubin Zheng is a seasoned engineer with over 20 years of experience in ICT and digital transformation. She focuses on building real-world AI systems, including agent runtime design, multi-agent coordination, and production deployment. She is an active contributor to developer and open-source communities, and a frequent speaker at cloud-native and AI events.