Manufacturing environments demand precision, reliability, and deep domain expertise to manage complex production processes, equipment performance, quality, and operational efficiency. This talk explores how AI models, specialized agents, and a robust execution harness can work together to address manufacturing-specific challenges. From root cause analysis and process optimization to equipment troubleshooting, quality assurance, and production planning, we examine how agentic AI can leverage engineering knowledge, manufacturing data, and operational tools to execute specialized tasks. The focus is on moving beyond general-purpose AI toward reliable, domain-aware agentic systems that deliver measurable improvements in manufacturing productivity, quality, and operational performance. The real opportunity isn't simply to make AI smarter. It's to make AI useful, dependable, and actionable in the environments where real work happens.