Jianqiang Guo
Head of Enterprise AI Agent Platform Product, 360 Group
Jianqiang Guo is Product Director at 360 Technology Group Co., Ltd. and Head of its Enterprise AI Agent Platform. He has over ten years of experience in B2B product management and platform development, with a long-standing focus on enterprise software, cybersecurity, and AI product planning and implementation. He currently focuses on building enterprise-grade AI Agent platforms, leading product design and practical implementation across agent development, runtime, evaluation, observability, the MCP/Skill ecosystem, context engineering, AgentOps, security governance, and private deployment and modular delivery. He is driving the evolution of the platform from an Agent Builder to an AI-native Agent Platform and continues to work with large enterprise customers on AI use-case planning and engineering implementation.
Topic
From Agent Builder to Agent Runtime: How Enterprise AI Agent Platforms Evolve from Building Agents to Delivering Real Business Outcomes
Enterprise AI agent platforms are evolving from simply building agents to enabling them to reliably accomplish tasks in real-world business environments. In practical deployments, models, prompts, knowledge bases, and tools are only the starting point. When agents operate across multiple systems, execute long-running workflows, and handle high-risk business processes, context management, MCP tool governance, identity and access control, execution constraints, exception recovery, human-in-the-loop mechanisms, continuous evaluation, and observability become critical to production readiness. Drawing on enterprise AI agent platform development and real-world customer deployments, this talk examines common challenges encountered in the evolution from Agent Builder to Agent Runtime. It also explores how Context, MCP, Harness, and AgentOps work together to establish a stable, trustworthy, and governable runtime environment for enterprise AI agents. Outline: 1. From Building Agents to Real-World Deployment: Key Challenges in Productionizing Enterprise AI Agents 2. Common Runtime Challenges: Context, Tool Calling, Long-Running Tasks, Identity and Access Control, and Security 3. Core Engineering Practices: Context, MCP Governance, Agent Harness, and AgentOps 4. Platform Evolution: From Individual Agents to Business Execution Environments and Production-Grade Governance
