Appointment reminders and no-show prediction
Your clinic loses roughly a fifth of its appointment slots to no-shows. Design an agent that manages reminders, predicts which patients are likely to miss, and offers freed slots to people on the waitlist — remembering that everything it touches is protected health information. Walk me through what patient data it needs versus what it must never see, how you'd roll it out safely and confirm the predictions actually hold, and give me the one-minute version for the practice director: problem, solution, payoff.
plan_no_show_actions(no_show_counts: list[int], past_appointment_counts: list[int], booking_lead_days: list[int], confirmed: list[bool]) → list[str][[2,0,1],[10,4,2],[7,45,14],[false,false,true]]out["extra_reminder","reminder","reminder"][[4],[8],[3],[false]]out["propose_release"][[],[],[],[]]out[]Turn the fuzzy goal into a bounded agent task. Say what the agent reads, what it may do on its own versus draft for a human, and what it must never touch. Then earn the trust: how you’d verify it before rollout, and the pitch (the problem, what you built, why it matters).
Agent build: design the agent in plain language (or narrate it). The coach grades the decomposition, the guardrails, the verification plan, and the pitch.