Residency Scheduling for Emergency Medicine Programs

Emergency medicine programs run around-the-clock shift patterns with night float, jeopardy, and weekend coverage requirements that interact across block periods. Thrawn's Scheduling Programming Language encodes all of these constraints simultaneously and generates a single, coherent schedule that satisfies coverage, ACGME duty hour rules, and fairness requirements at once — rather than solving each constraint separately and patching conflicts afterward.

Yes. ACGME duty hour rules — including the 80-hour weekly limit, required rest periods, and shift length caps — are built directly into Thrawn's optimization engine. Violations are prevented at the point of schedule generation, not flagged after the schedule is already built. This means your program director receives a finished schedule that is already compliant, with no manual review of duty hours required.

Thrawn produces Block, Call, Clinic, and Attending schedules for emergency medicine programs. All four schedule types are optimized simultaneously as one interconnected system, which prevents the common problem where updating one schedule type creates conflicts in another. Programs send their constraints and receive finished schedules for review.

Yes. Unlike Amion, which publishes and displays schedules but does not create or optimize them, or spreadsheets that require hundreds of hours of manual work with no compliance checks, Thrawn delivers fully optimized, ACGME-compliant schedules as a done-for-you managed service. Your program provides constraints; Thrawn produces the finished schedule.

Thrawn uses a mathematical fairness engine that distributes desirable and undesirable assignments — nights, weekends, holidays, and coveted rotations — with provable equity across all residents in your EM program. Because the distribution is produced by mathematical optimization rather than manual judgment, the process is transparent and defensible, which reduces scheduling complaints and removes the perception of bias.