Azaz Vepari07 Mins read
Probook Raised $40M to Automate Dispatch — Should ServiceTitan Worry?
Andreessen Horowitz and Sequoia put $40M behind Probook, which argues dispatch — not the phone — is where an AI operating system for home services has to start. The claim, the customer numbers, and who this is actually for.
In this article10 sections
Andreessen Horowitz and Sequoia have put $40 million behind Probook, a company arguing that dispatch — not the phone — is where an AI operating system for home services has to begin.
That argument is worth more attention than the cheque. Most AI money in the trades has gone to answering missed calls. Probook is claiming the harder job: deciding which technician goes to which job, and then owning everything that decision touches.
Its customer pages carry the kind of numbers that make competitors read twice. One HVAC, plumbing and electrical operator says it went from 22 dispatchers to 10. Another says 95% of daily dispatch decisions now run through the platform.
Those numbers are published by Probook. They are not independently audited, and this piece treats them accordingly — as evidence that large contractors are running the thing, not as a forecast for your shop.
The funding, precisely
| Item | Detail |
|---|---|
| Announced | 23 June 2026 |
| Series A | $34 million, led by Andreessen Horowitz |
| Seed | $6 million, led by Sequoia Capital |
| Total | $40 million |
| Note | Sequoia also participated in the Series A |
| Founders | George Eliadis (CEO), Lewis Zhang (CTO), Ben Cervantez |
| Based | New York |
This part is not a vendor claim. Two of the most recognisable firms in venture confirmed it, and the trade press carried it. a16z’s David Haber framed the thesis in one line: “Dispatch is the nerve center of every home service business.”
For a company whose focus is dispatch inside HVAC, plumbing and electrical businesses, $40 million is a serious position — and it is the reason the strategic question below is worth asking at all.
The argument: why dispatch first
Probook’s reasoning runs roughly like this.
Home service companies spent the last few years buying separate AI tools — one for inbound calls, one for web leads, one for outbound follow-up, one for customer messaging. Those tools did not share the same operational context. So the office ended up managing more vendors and more exceptions than before, which is the opposite of what was sold.
Dispatch, the argument goes, has to come first, because dispatch is where the customer promise, the right technician, the schedule, the expected ticket and the customer update all meet. Own that, and the surrounding workflows can be run from the same context rather than bolted alongside it.
In practice Probook says it handles intake, cleans the booking data, communicates with customers, works open estimates and memberships, and reshuffles the dispatch board using historical information — with humans managing exceptions while the AI takes the repetitive decisions and conversations around the board.
This is the part I think is genuinely right. Dispatch is not a map, and it is not route optimisation. The closest technician is very often the wrong technician. The right call can depend on skill, callback risk, expected ticket, membership status, promise window, equipment history, parts on the van, overtime exposure — and on what happens to the rest of the board once one emergency is inserted into it.
That is a substantially harder problem than answering a phone. It is also a more defensible one, which is presumably what the investors bought.
The customer numbers, and whose they are
Probook publishes outcome pages for several operators. Here is what two of them state.
Del-Air
An HVAC, plumbing and electrical company with 220+ service technicians across 8 Florida locations.
| Metric | Published figure |
|---|---|
| Dispatchers | 22 before, 10 after |
| Call assignment automated | 90% |
| Monthly inbound calls handled by AI | 16,000 |
| Containment rate | 74% |
| Tech-to-dispatch ratio | More than doubled |
Del-Air’s VP of Sales, Steve Ramthun, says Probook “was the only AI company we trusted to run our dispatch function”. Its Manager of Performance and Compliance, Terrence Fraser, describes the change in headcount terms: “The process used to require three people, from taking the call to placing the job to dispatching it.”
Peterman Brothers
200 technicians across 11 markets.
| Metric | Published figure |
|---|---|
| Dispatch decisions made by Probook | 95% |
| Non-dispatch inbound calls handled, one month | 2,500 |
| Jobs booked per month with no human involvement | 414 |
| HVAC average ticket, first 10 months year on year | +22% |
CEO Chad Peterman: “With Probook, we’ve centralised dispatch across 11 markets and 200 technicians without adding overhead.”
How to read all of that
These come from Probook’s own website and customer marketing. They are not independently audited comparisons.
What is missing is the same list that is always missing: the full labour baseline, the software and implementation cost, the job mix, seasonality, and every other operational change made during the same period. A company that consolidates dispatch across eight locations is usually changing several things at once, and the dispatcher reduction is being credited to the software.
One small illustration of why these figures deserve a light grip. The June funding release quoted Chad Peterman describing “11 markets and 180 technicians”; the customer page now says 200. Nothing improper about that — companies grow, and marketing pages get updated. It is simply a reminder that these are living marketing numbers rather than an audited filing, and that the version you read depends on the day you read it.
A note against our companion video. The video says Peterman has “more than 200 technicians”. Probook’s page says 200. The article is the record.
Treat all of it as evidence that operators at real scale are running the platform. That is a meaningful claim on its own — and it is a different claim from “you will halve your dispatch team”.
So should ServiceTitan worry?
Not in the way the headline implies — at least not on the public material available.
ServiceTitan still holds the system-of-record advantage: customers, jobs, price books, invoices, reporting, accounting workflow, and its own expanding AI stack. Probook’s current story is that it runs the customer experience and dispatch layer around an operation like that, not that it replaces the database underneath.
But there is a real strategic risk in that arrangement, and it is worth naming plainly:
The system employees use to make daily decisions can become more important than the database underneath it.
That is how control layers work. If dispatchers, CSRs and managers spend their day in Probook’s board while the platform of record quietly becomes a place data is written to, the commercial relationship shifts over time even though no migration ever happened. Whether that plays out depends on execution rather than on the size of a funding round.
For where the incumbents currently sit on dispatch specifically, our HVAC dispatch software comparison covers what ServiceTitan, Housecall Pro, Jobber, FieldCamp and the rest actually do on the board today, with pricing.
The catch: this is not an app install
Probook talks openly about deploying in person, learning the business, configuring complex rules and staying involved with the frontline team.
That is very likely why it can carry operators of this size — dispatch logic is idiosyncratic, and rules that work in eight Florida locations will not transfer unchanged to eleven markets in Indiana. Rick Rogers, Del-Air’s CEO, credits exactly that: “They took the time to come in, learn our business, and adapt their product to our strategy.”
It also means this is not a five-minute install for a three-truck company.
Pricing is not published on the pages we reviewed. No subscription figure, no implementation figure, no usage terms. So there is no ROI arithmetic available from the outside — you would need the full implementation and subscription quote in writing before any of these percentages mean anything financially.
Who should actually care
A large, multi-location operator with a heavy call centre and centralised dispatch should study this closely. That is the shape of every customer Probook publishes, and centralising dispatch across locations is where the dispatcher-reduction maths comes from.
A smaller shop with five technicians and one experienced dispatcher should not assume venture funding makes a platform a fit. At that size your dispatcher already holds the context the AI has to be taught, and complexity, data quality, change management and human oversight remain your problem either way.
The one-week test I would run
If I were evaluating Probook, I would give it one week of real historical calls and dispatch decisions — not a demo dataset — and ask it four questions:
- Why did each technician get each job?
- What constraints did it hit — capacity, skill, overtime, promise window?
- What did it change on the board, and what did that do to the rest of the day?
- What can a human override, and how quickly?
If the answers are legible and the board genuinely gets better without the heat and cleanup that follows a bad dispatch day, this is real. If the answer is a black-box revenue promise, keep looking.
Frequently asked questions
What is Probook?
An AI platform for home service operators that centres on dispatch — matching the right technician to the right job using historical shop data — and extends into intake, customer messaging and outbound workflows. Founded by George Eliadis, Lewis Zhang and Ben Cervantez, based in New York.
How much did Probook raise, and from whom?
$40 million total, announced 23 June 2026: a $34 million Series A led by Andreessen Horowitz and a $6 million seed led by Sequoia Capital, with Sequoia also joining the Series A.
What does Probook cost?
Not published. The pages we reviewed carry no subscription price, implementation fee or usage terms. Get all of it in writing before doing any ROI calculation from the percentages on its customer pages.
Is Probook a ServiceTitan replacement?
Not on the public material. ServiceTitan remains the system of record — customers, jobs, price books, invoices, accounting. Probook’s claim is to run the dispatch and customer-experience layer around that. The strategic risk to incumbents is layering, not migration.
Are Probook’s customer results independently verified?
No. The figures on this page come from Probook’s own customer-outcome pages and are not audited. They lack the labour baseline, cost, job mix and seasonality context needed to reproduce them. The funding figure is the exception — that one is independently reported.
Is Probook suitable for a small contractor?
Its published customers are large multi-location operators, and it describes an in-person implementation involving rule configuration and frontline training. A three-truck shop is not the profile these results come from.
The verdict
Probook is one of the most important AI-in-the-trades companies to watch in 2026 — not because of the $40 million, but because it is attacking the part of the business where customer experience and technician revenue actually meet.
The winner in this story is not a product. It is the idea of AI as an operating layer rather than another point tool — and if that idea holds, it reframes what every vendor in this category is selling.
Just remember which numbers are which. The funding is confirmed. The results are the vendor’s.
Sources
Sponsored
Ready to upgrade your dispatch?
FieldCamp uses AI to optimize routes, auto-assign jobs, and eliminate scheduling chaos — so you can run more jobs with fewer headaches.
Tested and written by
Azaz Vepari
Editor, Modern Field Service
Listen to this review
The same comparison as a podcast episode, for the drive between jobs.
Tool profiles
Pricing, best fit and limitations for each tool in this review, one page each.
