Hey Product Hunt 👋 Founder of Crisphive here. Let me tell you about the 45 minutes that made me build this. Every field service company lives the same morning: an emergency call comes in at 9am. A tech calls in sick. A job runs long. And suddenly a dispatcher is standing at a whiteboard — or worse, a drag-and-drop calendar — manually reshuffling the entire day. Calling customers. Re-routing crews. Guessing at travel times. Forty-five minutes later there's a "new" schedule that's already wrong, three customers got bumped without anyone noticing an SLA, and two techs are sitting in trucks burning windshield time waiting to be told where to go. The tools they're using aren't broken — they're just calendars. And a calendar can't answer the only question that matters: if this one job moves, what else has to move? That's a dependency problem, not a calendar problem. So we built Crisphive around it: 🔧 Real workdays as hard constraints. Travel time, mobilization/demobilization, breaks, supply runs, crew qualifications — modelled as constraints a solver must satisfy, not suggestions it can ignore. Every schedule it produces is one your team can physically execute. ⚡ Cascade rescheduling in under 3 seconds. When something slips, our deterministic solver resolves every affected downstream job — priorities and SLAs preserved, every decision logged and auditable. The 45-minute reshuffle becomes a non-event. 🤖 And no, the scheduler is not an LLM. Your AI agents can talk to Crisphive through our MCP server, but the actual scheduling decisions come from a deterministic solver. Same inputs, same schedule, every time. We think that's the only honest way to do AI scheduling for real crews — happy to debate this in the comments. What that's worth: dispatchers stop firefighting, techs spend hours on wrenches instead of windshields, and customers stop getting the "we need to reschedule" call a day late. You don't have to rip anything out — keep your FSM or CRM as the system of record and plug Crisphive in via API, SDK, or MCP. Or use it standalone. We charge per job, not per seat, and there's a free tier. 🧪 Try it right now, no demo call: the sandbox is open — break a schedule and watch the cascade fix it. 📊 PH-exclusive: we'll run a free shadow schedule audit on your real operation — we replay your past schedules against our solver and show you exactly what manual reshuffling cost you. I'll be here all day. Ask me anything about constraint solvers, why we're contrarian on LLMs-as-schedulers, or tell me your worst dispatch horror story — genuinely want to hear those.

Crisphive

constraint-based AI scheduling infrastructure

constraint-based AI scheduling infrastructure