Conference Description
Key Takeaways
- Technical summit dedicated to agentic AI systems, Model Context Protocol (MCP), and agent infrastructure
- Addresses production challenges including orchestration, security, observability, and scalability
- Designed for developers, platform engineers, architects, and enterprise teams building agent-based systems
- Emphasis on open-source development, cloud-native platforms, and emerging governance standards
- In-person event at the Shanghai International Convention Center featuring technical sessions and case studies
Introduction
AGNTCon + MCPCon China is a technical summit bringing together practitioners working on agentic AI systems and the protocols that underpin them. The event targets developers, researchers, platform builders, and enterprise engineering teams seeking to advance the reliable deployment of AI agents at production scale. With agentic architectures rapidly moving from experimental projects to operational systems, the summit addresses the growing need for shared knowledge around orchestration, security, and infrastructure standards.
The convergence of large language models with autonomous agent capabilities has created significant demand for robust protocols and tooling. Model Context Protocol (MCP) has emerged as a focal point in this space, providing a standardised approach for agents to interact with external systems and data sources. AGNTCon + MCPCon China positions itself at this intersection, offering deep technical content for those building the foundational layers of agentic infrastructure.
About This Event
Held at the Shanghai International Convention Center, AGNTCon + MCPCon China combines keynote presentations, technical sessions, and real-world case studies. The programme emphasises hands-on, practitioner-focused content rather than high-level overviews, reflecting the maturity of attendees expected to participate. Co-located events and certification opportunities extend the learning experience beyond the main conference sessions.
The summit serves as a meeting point for both local and international communities working on agentic systems. This cross-border dimension is particularly relevant given the parallel but sometimes divergent development of AI agent ecosystems in different regions. By facilitating direct exchange between these communities, the event aims to accelerate the establishment of interoperable standards and shared best practices.
Technical Focus Areas
The summit’s technical programme centres on the agentic stack—the layered architecture required to build, deploy, and manage AI agents in production environments. This encompasses agent protocols like MCP, which define how agents communicate with tools and retrieve contextual information, as well as the orchestration frameworks that coordinate multi-agent workflows.
Infrastructure considerations feature prominently, including cloud-native deployment patterns, observability tooling for monitoring agent behaviour, and security frameworks designed for autonomous systems. Unlike traditional software, agentic systems introduce unique challenges around unpredictable execution paths and dynamic tool usage, requiring specialised approaches to logging, debugging, and access control.
Governance and standards represent another significant thread. As organisations move agents into production, questions around accountability, auditability, and compliance become pressing. The summit addresses these concerns through sessions on emerging governance frameworks and the role of open-source communities in establishing industry norms.
Industry Context
The timing of AGNTCon + MCPCon China reflects a broader industry shift. Agentic AI has moved beyond proof-of-concept demonstrations into enterprise deployment, creating urgent demand for production-grade infrastructure. Organisations are discovering that building reliable agent systems requires more than capable language models—it demands robust protocols, careful orchestration, and comprehensive observability.
Open-source development plays a central role in this evolution. Many of the foundational tools and protocols for agentic systems have emerged from community-driven projects, and the summit reflects this by engaging project maintainers and foundation leaders alongside enterprise practitioners. This blend of perspectives helps bridge the gap between cutting-edge research and operational reality.
Who Should Attend
The summit is designed for technical practitioners directly involved in building or operating agentic systems. This includes open-source developers contributing to agent frameworks, platform engineers designing infrastructure for agent deployment, and enterprise architects evaluating how agentic capabilities fit within existing technology stacks. Researchers and academics working on agent behaviour, safety, or protocol design will also find relevant content.
Decision-makers responsible for AI strategy within their organisations may benefit from the event’s focus on real-world implementation challenges and emerging best practices. The technical depth assumes familiarity with AI systems and cloud-native development, making it most suitable for those with hands-on experience in these domains.
Ecosystem and Collaboration
Beyond formal sessions, AGNTCon + MCPCon China emphasises ecosystem development and community building. The event provides opportunities for practitioners to connect with peers facing similar challenges, share implementation experiences, and explore potential collaborations. For organisations building products or platforms in the agentic AI space, the summit offers visibility within a concentrated audience of technical influencers and early adopters.