AI systems show potential for transformative impact in a wide range of application areas, from customer support and system monitoring to software development. Successful integration of AI tools can, however, require fundamental changes to how an organization runs. In this class, we will discuss how best to address operational challenges that arise in AI-centered organizations, and seek to understand when and where AI tools have the most potential for impact. We will also discuss design of agentic systems, resourcing and routing for AI workflows, complementarities between AI and non-AI systems, human-in-the-loop designs, AI evaluation, and safety. This class will include three components: first, lectures on conceptual and technical underpinnings of AI systems, their implementation, and their evaluation; second, hands-on coding and data work culminating in a system-design project; and third, guest lectures from industry about AI implementation and use cases.
2 units · GSB Letter Graded
AI systems show potential for transformative impact in a wide range of application areas, from customer support and system monitoring to software development. Successful integration of AI tools can, however, require fundamental changes to how an organization runs. In this class, we will discuss how best to address operational challenges that arise in AI-centered organizations, and seek to understand when and where AI tools have the most potential for impact. We will also discuss design of agentic systems, resourcing and routing for AI workflows, complementarities between AI and non-AI systems, human-in-the-loop designs, AI evaluation, and safety. This class will include three components: first, lectures on conceptual and technical underpinnings of AI systems, their implementation, and their evaluation; second, hands-on coding and data work culminating in a system-design project; and third, guest lectures from industry about AI implementation and use cases.
Offered in Winter 2027 at Stanford University.