ShiftSense
Built at Meta Global AI Developer Hackathon · Oct 26, 2026 · Remote

A charge nurse spends about two hours building each fortnight's roster by hand, then reworks it every time someone calls in sick. ShiftSense generates the roster in seconds with a CP-SAT optimizer that ranks what matters: cover every shift, change as little as possible, respect contracts, honour requests, and share nights and weekends fairly. Nurses message the app in Hindi or English, by text or voice note. Muse Spark understands the request, and an agent finds replacements and contacts them one by one. Muse never edits the roster itself. Every change is re-solved by the optimizer and published only when the charge nurse approves it.