AI dispatcher for home services: scheduling jobs and technicians

Every home services firm has a person, or a small team, who holds the day together: the dispatcher. They take the first call, work out whether it is a blocked drain or a failed boiler, decide who is closest and qualified, and then rework the whole board when a technician runs late. ServiceTitan surveyed 500 dispatchers and reported that half rated their stress at 7 or higher out of 10, and 74% believed their shops needed more staff, according to its Dispatch Pro webinar recap.

That pressure is why an AI dispatcher for home services has moved from curiosity to a serious buying question for plumbers, electricians, heating engineers and garage door firms. This guide explains what an AI dispatcher actually does, where it helps, where it still needs a human, and how to connect it to the field service software you already use.

Quick answer

An AI dispatcher for home services takes job requests, works out urgency and skills needed, matches the job to the right technician by location and availability, and reshuffles the day when plans change. It suggests or makes assignments inside your field service software, while a human dispatcher keeps control of emergencies, exceptions and anything unusual.

What an AI dispatcher does in practice

The phrase covers several jobs that used to sit with one person. It helps to separate them, because most firms need some of these and not all of them.

Salesforce describes the same set of tasks in its field service AI guide. It says AI scheduling weighs job location, technician skills, job urgency, travel time, parts inventory and customer preferences, and that AI assists dispatchers with sudden changes such as cancellations or a technician calling in sick by flagging the exception and recommending how to rework the schedule.

Notice what that list does not include. It does not include deciding that an engineer should attend a gas leak without a person checking it, or promising a customer something the firm cannot deliver. The useful design principle is that the AI does the repetitive weighing of options and the human keeps the judgement calls.

Where the time goes without it

Most owners underestimate how much of a dispatcher's day is lookup and reshuffling. A job comes in. Someone checks which technicians hold the right qualification, then checks where each one is, then checks whether moving a booking will upset a customer who was already promised a slot. Then a van gets stuck in traffic and the whole exercise repeats.

The front of that process is just as exposed. Invoca reported in its article on what missed sales calls cost home services businesses that 27% of calls to home services businesses go unanswered, and that fewer than 3% of callers sent to voicemail leave a message. Invoca is a call analytics vendor and the article cites its own research without a year, so treat the figure as an indication rather than a benchmark. It still matches what most owners see: a call missed while a team is on site is often a job lost to the next firm on the search results.

This is where the receptionist role and the dispatcher role overlap. If you want the intake side covered in depth, our guide to an AI receptionist explains how call answering works. The dispatcher layer sits behind it and decides what happens to the job once it is captured.

How the major field service platforms handle dispatch

Before you build or buy anything, check what your existing software already does. Several platforms now ship dispatch intelligence, and a custom layer only makes sense where it fills a real gap.

ServiceTitan Dispatch Pro

ServiceTitan's Dispatch Pro page says the tool uses Titan Intelligence to suggest which technician should take each job, considering skills, performance, location and drive time. Dispatchers can choose between on-demand suggestions with Dispatch Assist and fuller automation, set maximum drive times, and prioritise drive time, average ticket size or a mix.

The company also publishes results from select customers measured over six months before and after implementation, including a 14% boost in average ticket size and a 2X increase in dispatcher efficiency. ServiceTitan notes that individual results vary, and the page gives two different drive time reductions (6% on the main page and 10% in the FAQ), so use these as vendor claims, not guarantees.

The most useful detail for anyone worried about losing control is in the webinar recap. ServiceTitan describes Dispatch Pro as a co-pilot that does not replace dispatchers. Settings let you lock jobs shortly before the start time, keep urgent recalls and warranty calls under manual control, and make a job technician-specific so it is never reshuffled. The recap also says the predictions work well after roughly three to four months of data, which is a reminder that these systems learn from your history rather than arriving finished.

ServiceTitan's summer 2025 product update goes further on the phone side. It describes AI virtual agents in early access that answer calls and book jobs using real-time capacity data, with live escalation to a human. The same update lists business-unit-based skills so that, for example, a plumber is not sent to an HVAC job.

Jobber

Jobber's feature list on its UK pricing page covers scheduling, crew scheduling, a Find a Time feature that recommends slots based on team availability and travel times, route optimisation that adjusts when plans change, online booking, and an AI Receptionist that answers calls and texts and can book jobs into the calendar. It also lists API access and a Zapier connection to more than 7,000 apps. For smaller teams already on Jobber, that may be enough to start with.

Salesforce Field Service

Salesforce positions a Dispatcher Agent for scheduling and exception handling and a Scheduling Agent that books appointments through a customer conversation, matching skills, tools and travel time, again according to its field service AI guide. The guide contains no statistics specific to AI dispatch, so there is nothing numerical to hold it to.

What the evidence says about results

Be careful with the numbers in this area. Almost all published results come from the vendors selling the software, usually from their best customers.

One example is ServiceTitan's A1 Garage Door case study. A1 reported that its ratio of technicians per dispatcher rose from roughly 10 to 12 up to 20 to 1, and its COO said the firm had added 150 technicians without adding a dispatcher. The firm also reported that lead conversion rose from around 9% to 9.5% to around 11.5% to 12%. These are the customer's own statements on the vendor's site. They show what a large, well-run operation achieved, and they say little about a six-van plumbing firm.

A fair reading is that the gains come from two places. First, dispatcher time is freed, so the same people can supervise more technicians and spend longer on quality checks. Second, better matching and tighter routing put more productive jobs in each day. Neither is magic. Both depend on clean data about technician skills, job types and service areas.

Where an AI dispatcher still needs a human

This is the part vendor brochures skip. We would keep a person in charge of the following.

Our view is that the right target is a dispatcher who handles twice the volume, not a firm with no dispatcher. A system that quietly runs the whole board with nobody watching will eventually make an expensive mistake on a Friday afternoon.

Data protection and automated decisions

Dispatch decisions mostly concern technicians and jobs, not people's rights, so most scheduling sits comfortably outside the strictest rules. Intake is different, because you are handling customers' names, addresses, voice recordings and sometimes details about vulnerability or health.

For UK firms, the ICO's automated decision-making guidance uses the definition in Article 22A of the UK GDPR: a decision based solely on automated processing, including profiling, with a legal or similarly significant effect. Choosing which van attends a leaking tap rarely meets that bar. Declining a customer, applying a surcharge or flagging someone as a bad risk might. The ICO's page on what else to consider recommends a data protection impact assessment for automated decision-making and says that monitoring a system's statistical accuracy is not the same as human intervention.

The government's summary of the Data (Use and Access) Act 2025 says the Act creates a more permissive framework for solely automated significant decisions, but requires safeguards: telling people, letting them make representations and challenge the decision, and letting them request human intervention. The ICO page we read notes that its guidance is a draft under consultation, so check the current version before you rely on it. Our guide to GDPR and custom AI agents in the UK and EU goes through the practical steps. This is general information, not legal advice, so take advice for your own situation.

How to connect an AI dispatcher to your field service software

Most owners do not need to replace their system. The sensible route is a layer that reads from and writes to the tools you already run.

Jobber's developer documentation shows how this works in practice. It is a GraphQL API with a single endpoint, OAuth 2.0 bearer tokens, queries to read data and mutations to create or change it. The page notes that mutations return user errors when a request is rejected, and that some errors can arrive with an HTTP 200 status. That detail matters for reliability: an integration that assumes a 200 response means success will quietly drop jobs.

The same principles apply whichever platform you use. Our article on how AI agents integrate with existing SMB tools sets out the pattern in more detail. In short:

How Klevere approaches an AI dispatcher

Klevere has deployed 500+ AI agents across 50+ projects and 12 industries, and we keep a 98% client retention rate. For home services we do not start with a model. We start by watching how dispatch really works in your firm, because the written process and the real one rarely match.

We map the job types you actually sell, the skills and licences each technician holds, your service areas and your rules for emergencies. We then check what your field service platform already offers, because buying nothing is sometimes the right answer. Where there is a gap, we build an agent that works inside your existing software through its API, rather than asking you to switch tools.

Our operations agent is the usual starting point for scheduling and re-planning work, and it pairs with intake from phone, text and web. If phone capture is the weak spot, we connect the dispatcher to an AI voice agent and keep the human escalation route in place.

We also tell clients when the project is not worth doing. A three-person firm with a quiet diary may gain more from answering every call than from automated technician matching.

How to start without a big project

A sensible first month looks like this.

Only after that should you let the agent book anything without approval, and only for the categories where its suggestions were right almost every time.

Frequently asked questions

What is an AI dispatcher for home services?

It is software that takes job requests, judges urgency, matches jobs to technicians by skills, location and availability, and rebuilds the schedule when things change. It works inside your field service software and hands difficult or urgent cases to a person. It supports a dispatcher rather than replacing one.

Will an AI dispatcher replace my human dispatcher?

Not in a well-run setup. ServiceTitan describes its own Dispatch Pro as a co-pilot that does not replace dispatchers, and lets you keep recalls and warranty work under manual control. The realistic gain is one dispatcher supervising more technicians and spending less time on repetitive reshuffling.

Does an AI dispatcher work with ServiceTitan or Jobber?

Yes. ServiceTitan has built-in dispatch intelligence, and Jobber offers scheduling, routing and an open GraphQL API for custom connections. A custom agent can work alongside either by reading jobs and technician data and writing bookings back, usually starting in a suggestion-only mode.

How long does it take to see results?

It depends on your data quality and job volume. ServiceTitan says its predictions work well after about three to four months of data, so expect a learning period. Running in suggestion mode first lets you judge accuracy within weeks, before any automation is switched on.

Is an AI dispatcher compliant with UK GDPR?

It can be, if you handle personal data lawfully, run a data protection impact assessment and keep meaningful human involvement in any decision with a significant effect on a person. Routine technician scheduling rarely falls into that category. Take legal advice for your own circumstances.

Can an AI dispatcher handle emergency call-outs?

It can recognise keywords and flag urgency, but we recommend a person confirms safety-critical emergencies such as gas or electrical faults. The agent should escalate immediately with the details captured, so the human spends their time deciding rather than gathering information.

If you want to know whether an AI dispatcher would pay back in your firm, or whether your current software already covers it, book a free AI audit and we will map your dispatch process and tell you plainly where automation helps and where it does not.