Why We Rebuilt the Scheduling Engine from Scratch

Today, we are sharing a look behind the curtain at one of the most significant engineering efforts in FieldPulse history: a complete rebuild of our scheduling engine from the ground up. This project has been eighteen months in the making, and it is now live for all customers on FieldPulse 3.3.

The Limitations We Hit

Our original scheduling system served us well through five years of growth. However, as our customer base expanded to include teams of 500+ technicians and complex multi-day projects, we began to encounter hard limits.

The legacy engine struggled with:

• Concurrent drag-and-drop operations across large dispatch boards
• Real-time conflict detection when multiple dispatchers edited the same schedule
• Scalable performance as daily job volumes exceeded 10,000 per tenant
• Integration flexibility with third-party calendar and ERP systems

We reached a point where incremental improvements would not deliver the reliability and responsiveness your operations require.

What We Built Instead

The new scheduling engine is built on an event-sourced architecture with conflict-free replicated data types (CRDTs). In practical terms, this means:

• Real-time synchronization across all connected devices without lock contention
• Intelligent conflict resolution when two dispatchers modify the same technician's schedule simultaneously
• Sub-second load times for dispatch boards with 500+ active jobs
• Native support for complex scheduling rules including skill matching, union requirements, and travel time optimization

We also took this opportunity to implement the foundation for route optimization, which launched in FieldPulse 3.2 and now operates seamlessly within this new architecture.

What This Means for Your Team

The most visible change is speed. Dispatch boards that previously took 8–12 seconds to load now appear in under one second. Drag-and-drop scheduling feels instantaneous, even during peak morning dispatch hours.

Less visible but equally important: the new engine eliminates an entire class of synchronization bugs that could cause jobs to temporarily disappear from a dispatcher's view or technician assignments to revert unexpectedly. If you have encountered either issue in the past, you will find them resolved in 3.3.

Looking Forward

This rebuild unlocks capabilities we have long wanted to deliver. In the coming quarters, you can expect:

• Automated scheduling suggestions based on historical patterns and real-time conditions
• Deeper integration with workforce management and payroll systems
• Advanced capacity planning tools for multi-week project scheduling

We recognize that a foundational change of this magnitude requires trust. Every scheduling operation in FieldPulse now runs through this engine, and we have invested heavily in observability and automated recovery to ensure it performs under all conditions.

If you have questions about the technical architecture or want to discuss how this impacts your specific deployment, I am available in the comments below.