Industry / Construction & trades

AI for the field, not the boardroom.

Your dispatch board is a whiteboard. Your invoicing is two days behind. Your best estimator is your owner and she is also running payroll on Sunday night. This is where AI can actually move a trades shop, and where it can't.

Book a Scoping Call See the readiness audit

Christopher Myers

2026 / Industry brief

Scene / 01

An HVAC shop · Commerce City, CO

It is 6:47 a.m. The owner is at the kitchen table with a coffee, looking at a dispatch whiteboard app on his phone. Twelve techs, 47 service calls queued, two emergency no-heat calls that came in overnight.

His office manager quits in three weeks. She has been the glue between dispatch, invoicing, and parts ordering for eight years. Nobody else knows how she sequences Tuesday routes so the boiler techs do not cross town twice.

The owner thinks about AI because his son-in-law keeps sending him articles. What he actually needs is not a tool. It is the office manager's knowledge, out of her head, and into a system that ships work to the right tech at the right time.

What's actually broken in the field.

Most trades software is not underpowered. It is under-configured. The office manager knows tricks the software does not.

Labor shortages are real. BLS projects a shortfall of roughly 500,000 skilled-trades workers year over year into 2027, and every owner feels it. But the most common bottleneck we see is not field labor. It is office throughput. One office manager typically supports 8 to 15 techs. When she is sick, the dispatch board freezes.

Second, estimating is the highest-leverage work in the shop and the most under-invested. The owner either does it (terrible opportunity cost) or a senior tech does (even worse, because now they are not billing). A fast, accurate estimate wins jobs. A slow one loses them to the competitor who answered the phone first.

Third, cash conversion cycle. Most trades shops invoice 48 to 72 hours after a job closes out. Every hour of delay is working capital tied up. Automating invoicing from the technician's in-truck notes, with a same-day drop to the customer's inbox, is one of the single highest-ROI moves in any service trade.

Sources referenced

BLS Occupational Employment Projections — Skilled-trades labor shortfall forecasts through 2027.
ServiceTitan State of the Trades (2024) — Vendor benchmark data on dispatch density, revenue per tech, and admin cost ratios.
IBISWorld HVAC / Plumbing industry reports — Margin compression and labor-cost trends for independent shops.

Five workflows AI changes for a trades shop.

These are tool-neutral scopes. Whether you run ServiceTitan, Housecall Pro, Jobber, or a handful of spreadsheets, the workflows underneath are the same. We pick tools to fit the shop, not the other way around.

Workflow 01

Answer the phone in 8 seconds, qualify the call, book the right slot.

What we'd build

A call-triage assistant on your main line that captures the lead, identifies emergency vs. routine, pulls customer history if they are a repeat, offers next-available slots based on dispatch density, and books straight into your field-service system. Your office stops choosing between answering calls and dispatching trucks.

Vendors we'd evaluate

ServiceTitan Voice
CallRail
Goodcall
Dialpad Ai

Vendor-neutral. No reseller margins.

Workflow 02

Turn the technician's truck notes into invoices before the tailgate closes.

What we'd build

A post-job workflow that takes the tech's voice notes or photos, assembles a clean invoice with parts and labor itemized, fires it to the customer for approval, and queues payment. Cash conversion drops from 72 hours to the same afternoon.

Vendors we'd evaluate

ServiceTitan
Housecall Pro
Jobber
Workiz

Vendor-neutral. No reseller margins.

Workflow 03

Produce a two-page estimate in an hour, not a week.

What we'd build

An estimating layer that takes site notes, photos, and your pricebook, drafts a scoped estimate with margins baked in, and spits out a professional proposal ready for your senior tech to review and send. The close-rate improvement is usually measurable inside a quarter.

Vendors we'd evaluate

Socotra Estimate
Knowify
Jobber Quotes + AI
STACK

Vendor-neutral. No reseller margins.

Workflow 04

Route the fleet like the office manager would.

What we'd build

A dispatch-optimization layer that takes your real constraints (tech skill, parts on truck, SLA windows, traffic), sequences tomorrow's routes, and adjusts intraday when emergencies land. Output is not a magic route. It is a draft schedule your dispatcher tunes instead of builds from scratch.

Vendors we'd evaluate

Geotab
Workwave Route Manager
OptimoRoute
Samsara AI

Vendor-neutral. No reseller margins.

Workflow 05

Win every review, answer every one-star, keep the pipeline warm.

What we'd build

A post-job review and reputation workflow. Customer gets a review ask at the right moment (after a successful job, by SMS, with a direct link). Negative reviews get drafted responses for the owner to approve. Warm leads who did not book get a scripted re-engagement in 30 days.

Vendors we'd evaluate

NiceJob
Podium
BirdEye
ServiceTitan Marketing Pro

Vendor-neutral. No reseller margins.

The independent-shop edge over the roll-ups.

“Private-equity HVAC roll-ups are buying your competitors. Their advantage is capital. Yours is speed and craft. AI widens both.”

The trades are being rolled up. You have seen it. A four-truck shop three cities over just got acquired by a holding company that now owns nineteen regional HVAC brands. The pitch they make is scale economics. What they lose in the process is the owner on the roof, the relationship with the customer, and the willingness to say no to work that doesn't fit.

You still have all three. The only thing the roll-up beats you on is back-office leverage. That is exactly the gap AI closes. A well-implemented dispatch layer, invoicing automation, and a real estimating flow gives your office-manager-of-one the throughput of a roll-up's regional office. The craft stays yours. The admin burden shrinks.

The math on timing is sharp. A shop that deploys two AI workflows in 2026 has a material cost advantage over a shop that doesn't, measured in recovered owner hours alone. Wait two years and the advantage is gone, because the roll-ups will have moved faster.

Mid-post · 30-minute scoping call

Want a 30-minute scoping call for your shop?

Bring the dispatch board, the invoicing cycle, and the piece of your week that hurts most. We will point at the workflow with the strongest first-sprint ROI and tell you if you are ready to run it yet.

Book the Call

Three things AI won't fix in the field.

These are the conversations we have on the truck, not in a pitch deck. Worth being honest about.

If the pricebook is broken.

AI estimating is only as good as the pricebook under it. If your parts and labor catalog is out of date or inconsistent between techs, the first sprint is fixing that, not layering AI on top of a mess.

If the techs do not use the app.

Every AI workflow we deploy in field services depends on the techs actually using whatever system we integrate with. If adoption on your current software is under 60%, we are going to tell you adoption is the problem and AI is a distraction.

If dispatch runs out of the owner's head.

A senior owner-dispatcher is often the single highest-value person in a small trades shop. We will not replace that judgment. We can build a system that drafts tomorrow's routes so the owner tunes them in 20 minutes instead of building them in two hours.

How we'd work with your shop.

01Diagnose

02Build

03Maintain

The Readiness Audit is two to four weeks. We ride along with a dispatcher, spend an afternoon in the truck with a senior tech, and sit with the office manager for a full workday. You walk out with a readiness score across eight dimensions, a prioritized workflow sequence, and rough ROI ranges on each.

The first sprint is almost always one of two: call handling and invoicing, or dispatch and estimating. We scope it in writing, run weekly demos with the owner and the office manager, and build inside whatever field-service system you already run. Training for the techs is part of the scope, not an add-on.

Managed is month-to-month. We monitor dispatch density, invoice-cycle time, and close rates. One to three new automations per quarter. Most shops see the retainer paid for by recovered owner hours inside the first 90 days.

Start with a readiness audit See pricing and scope

Questions owners actually ask.

The questions operators in this vertical actually ask on the first call.

01We run ServiceTitan / Jobber / Housecall Pro. Does that change the scope?

It changes the integration work, not the shape of the engagement. Each platform has different API depth. We pick workflows that play to your platform's strengths and patch the gaps with best-of-breed vendors where needed.

02Can AI really answer our phones? Our customers are mostly older homeowners.

Modern voice models are a lot closer to a real person than a traditional IVR. For emergency calls we route straight to a human. For routine booking and qualification, most customers do not notice, and we always offer a path to a live person. We measure satisfaction before and after.

03How do you handle techs who aren't comfortable with new tech?

By building around them rather than on top of them. Voice-first capture in the truck, minimal typing, no extra apps. The best way to lose a shop is to put a third app on a tech's phone. We try not to.

04What about estimates that depend on the owner walking the site?

AI does not replace the site walk. It replaces the four hours of writing that happen after the site walk. The owner shoots a few photos and voice notes. The estimating layer produces the polished proposal.

05Are you going to tell us to rip out our software?

Almost never. Most shops already own software with 70% of what they need. The work is configuration, integration, and adoption. Replacements are rare and we flag them only when there is no other path.

06Do you work with owner-only shops (1 to 5 techs)?

Below five techs, we usually point you at off-the-shelf tools (Jobber, Housecall Pro, Voicedrop) without a custom build. We will tell you so during the scoping call and refer rather than sell you a sprint you don't need yet.

End of post · Next step

Your shop doesn't run on strategy decks. It runs on Tuesday.

Thirty minutes, a scoping call. Bring the thing that is eating your Sundays. If AI is the right next move, we will scope it. If it isn't, we will tell you.

Book a 30-Minute Assessment See the readiness audit

What the 30 minutes delivers

01A short list of AI opportunities specific to your shop.
02A rough ROI range and a sense of which to build first.
03An honest answer: audit now, wait a quarter, or skip us.

Free · 30 minutes · No deck