This workshop examines how Posit Assistant in Positron can support analysis and reporting while keeping methodological decisions and code execution under human control. Participants explore a workflow in which an approved language model helps draft R or Python code for tables, listings, figures, and reports. The emphasis is on reviewing generated code, choosing the analytical approach, and approving execution within the local analysis environment. Examples span pharmaceutical research and development contexts, including clinical trials and pharmacokinetics. The session also addresses monitoring and optimizing token usage, helping participants understand how to incorporate AI assistance into a deliberate, supervised programming workflow.