Introducing Field Notes: Dispatches from the Front Line of Agentic Scheduling
Field Notes is Crisphive's working blog: what we learn building agentic scheduling for field teams — dispatch practice, AI integration guides, and the occasional war story from the solver.
Most content written for field operations falls into two categories: high-level sales brochures pitching vague efficiency gains, or generic advice aimed at people who have never spent an afternoon juggling callbacks, technician skill sets, and a late-running truck. Very little of it covers what actually happens when an automated system or AI agent meets a live dispatch board with hard real-world constraints.
We are launching Field Notes to document that exact operational reality. This blog is a place for practical working notes on how field operations scheduling functions behind the scenes, how AI agents can interact safely with live dispatch boards, and how to structure dispatch workflows without breaking commitments to customers.
What we'll write about
Dispatch practice. We will focus on the concrete mechanics dispatchers and operations managers navigate daily: balancing SLA windows against route density, managing mid-afternoon emergency job insertions, reducing unnecessary windshield time, and designing job workflows that technicians can follow on site.
Building with AI. We will publish technical tutorials on how AI agents evaluate scheduling rules, verify constraints, and interact with live dispatch data. This includes practical guides on using the Model Context Protocol (MCP) to let models like Claude or ChatGPT read and manage field schedules. Learn more about our MCP server interface for AI agents.
Engineering notes. We will share details on building deterministic, constraint-based solvers. When you factor in technician skills, working hours, map-drawn service areas, travel times, and planned time off, automated scheduling requires a predictable solver that yields identical, reliable outputs every time rather than an LLM guessing at open calendar slots. Explore how this logic operates on our solution page.
What we won't do
We will not publish manufactured case studies, corporate announcements, or stories built around hypothetical customer quotes. When we analyze operational trade-offs, we will outline the actual mechanisms—such as why a second truck roll to deliver a missing part costs far more than the margin on the job itself. We will share data only when we can stand behind it, and we will avoid single-click claims about complex dispatch problems.
Start here
If you are ready to begin, read our guide on connecting field operations software to Claude. To see how these connections work in practice with a complete workflow, follow our tutorial on building an AI dispatching agent with Claude and Crisphive MCP.
To evaluate how your own dispatch constraints and schedule rules are performing, request a free operational audit at /audit with Crisphive.
