AI for Trades: The Dispatch Setup

Connecting Botpress Workflows to Field Service Management (FSM) Software

For a plumbing or HVAC business, the "Leaky Bucket" isn't just a service call—it’s what happens to their revenue when a high-ticket emergency call goes to voicemail. As a developer who has spent a quarter-century architecting software solutions, I’ve seen that the most expensive lead is the one you already paid for but failed to capture.

The Triage Architecture

The heart of a professional trade bot isn't just "chatting"—it’s triage. Using the SkyAgent: Home Services logic, we separate the "I have a small leak next week" from the "My basement is currently a swimming pool".



1. The Intake Node: Collects the raw problem description from the user.
2. The AI Classifier: Uses a fine-tuned system prompt to identify keywords like "flooding," "no heat," or "gas smell".
3. The Conditional Fork: If workflow.is_emergency is true, the bot skips the pleasantries and moves straight to rapid data collection.

Bridging the Gap: Webhooks to FSM

A bot is only as good as the data it delivers. To make this a "turnkey" solution, we use Webhooks to push lead data into industry-standard tools like ServiceTitan or Housecall Pro.

• Data Payload: We package the customer_name , phone_number , and problem_desc into a JSON object.
• The Handoff: For emergency leads, the bot can trigger an immediate SMS alert to the technician on call, ensuring the "Speed to Lead" is measured in seconds, not hours.