Stellaris Ridge
AI agents booking home service jobs sounds like a headline from five years out.
It is already live infrastructure, running in production, at over 150 organizations.
If you run a local trades or home-service business, the buying journey you have been optimizing for — a human searches, a human compares, a human calls — is starting to get a middleman, and that middleman is software acting on the customer’s behalf.
This is not a warning about robots taking over.
It is a walkthrough of what is already built, what it means for how customers hire you next, and why the thing an agent transacts with is your data, not your homepage.
The old buying journey, and the one replacing it
The buying journey most local businesses still plan around looks like this: a homeowner has a problem, searches Google, reads a few sites or an AI summary, picks up the phone, and talks to a person.
A new version of that journey is being built right now by the largest platforms in tech, and it looks different.
You → Your Agent → Business Agent → Booked Job.
The homeowner asks their personal AI agent to handle it.
Their agent researches local options, compares specifics — price, availability, service area, reviews — and communicates directly with the business’s systems.
The business’s own agent, or its structured data, answers back with real availability and a real quote.
The job gets scheduled, sometimes before the homeowner has spoken to a human at all.
That is not a hypothetical flow.
Pieces of it are shipping right now, from three different directions at once.
Piece one: Gemini Spark, the personal agent doing the asking
Google announced Gemini Spark at Google I/O 2026 as its personal AI agent for paid Google AI Ultra subscribers.
Spark researches options, compares them, gets quotes, and schedules appointments on the user’s behalf.
Think of it as a homeowner’s personal assistant that never gets tired of calling around for quotes.
Today it is limited to Ultra subscribers, which sounds like a small slice of the market.
But every major platform shift starts with the paid early-adopter tier before it becomes the default experience, and the direction is unmistakable: fewer homeowners doing their own comparison shopping, more of it delegated to an agent.
Piece two: A2A, the protocol agents use to actually talk to each other
Google’s Agent-to-Agent protocol, known as A2A, launched in April 2025.
It has since moved to version 1.2 and is now governed by the Linux Foundation, with roughly 150 organizations running it in production.
That detail matters more than it sounds like it should.
A2A is not a lab experiment or a whitepaper.
It is a shared, standardized way for one AI agent — say, a customer’s personal assistant — to communicate and transact with another AI agent, such as a system representing a business.
150 organizations in production means the plumbing between agents already exists and is being used today, quietly, by companies that already see where the buying journey is headed.
Piece three: ARD, the way an agent finds and verifies your business at all
None of the above works if an agent cannot find your business or trust what it finds.
That is what the Agentic Resource Discovery spec, or ARD, solves.
Google, Microsoft, GoDaddy, Salesforce, and others published the ARD spec on June 17, 2026.
It defines a small manifest file — ai-catalog.json — that a business publishes on its own domain, the same way a robots.txt file has quietly sat on every website for decades telling search crawlers what they can access.
Alongside the manifest, ARD defines a registry API that AI agents use to discover and verify a business: what it offers, where it operates, how to get a real quote.
Put the three pieces together and the picture is complete.
A homeowner’s Gemini Spark agent researches HVAC companies, talks to each business’s agent or structured data over A2A, verifies who is legitimate and current through the ARD registry, and books the appointment.
The homeowner may check the calendar invite and not much else.
📞 Want your phone answered every time it rings?
Stellaris Ridge builds the AI so you don't have to think about it. VOX answers every call 24/7, sorts the emergency from the routine job, and books it on your calendar. You just see more jobs.
→ Talk to us here
Why your homepage isn’t what the agent is talking to
This is the part that should change how you think about your website.
An agent does not “visit” your site the way a human does, scrolling and reading a hero image and a call-to-action button.
It reads structured, verifiable facts: your services, your pricing, your service area, your availability, your credentials.
A gorgeous homepage with a vague “contact us for a free estimate” button gives a human something to look at and gives an agent almost nothing to act on.
The business that wins the agent-to-agent version of the buying journey is not the one with the prettiest site.
It is the one with the most complete, most verifiable, most current structured data — the kind an agent can check, trust, and transact against without a human in the loop.
That is a fundamentally different asset than a marketing website, and it is worth building deliberately rather than assuming your existing site already covers it.
We go deep on exactly what that asset looks like, layer by layer, in The Knowledge Catalog — the practical build guide for the structured data an agent actually reads.
What this means for a shop with 1 to 15 employees
None of this requires you to build your own AI agent or write a line of code.
It requires the same foundation every serious local business needs regardless of which protocol wins: accurate services, honest pricing, a complete and current business profile, and a clear way for anyone — human or agent — to book a real appointment.
The businesses that get skipped in this new buying journey will not be skipped because they did anything wrong.
They’ll be skipped because there was nothing verifiable for the agent to check.
The businesses that get chosen will be the ones whose data was simply there, correct, and current when the agent came looking.
That is the whole game, and it is a game a small owner-operated shop can win just as easily as a national chain, because completeness and accuracy do not scale with headcount.
📞 Want your phone answered every time it rings?
Stellaris Ridge builds the AI so you don't have to think about it. VOX answers every call 24/7, sorts the emergency from the routine job, and books it on your calendar. You just see more jobs.
→ Talk to us here
Where field service is already moving
This is not purely theoretical for the trades, either.
Field-service platforms are already building for an agent-ready world.
Jobber, one of the most widely used platforms among small home-service businesses, now publishes a dedicated information page aimed at answer engines and sells its own AI booking receptionist as part of the platform.
That is a strong signal that the shift described here is not a distant forecast from a Google keynote.
It is showing up in the everyday software tools local trades businesses already use to run their day.
The businesses paying attention now are the ones building the structured foundation before it becomes table stakes, the same way the businesses that took SEO seriously in 2010 had a decade-long head start over the ones who waited.
If you want the immediate, practical version of this shift — what changed in Google search results just this month and what to publish today — read Your Website Isn’t the Front Door Anymore, which walks through the zero-click search numbers driving all of this.
Frequently asked questions
What does it mean for AI agents to be booking home service jobs?
It means a customer’s personal AI assistant researches, compares, and books a service appointment on the customer’s behalf, communicating directly with a business’s own systems rather than the customer browsing websites and calling businesses one by one.
What is Google’s A2A protocol?
A2A, short for Agent-to-Agent, is a protocol Google launched in April 2025 that lets one AI agent communicate and transact with another AI agent using a shared, standardized format.
It is now at version 1.2, with roughly 150 organizations running it in production under Linux Foundation governance, meaning agent-to-agent commerce is live infrastructure today, not a future concept.
What is Gemini Spark?
Gemini Spark is Google’s personal AI agent, announced at Google I/O 2026, currently available to paid Google AI Ultra subscribers.
It researches options, compares businesses, gets quotes, and schedules appointments on the user’s behalf, acting as a digital assistant that handles the legwork of hiring a local service business.
What is the ARD ai-catalog.json spec?
ARD, or Agentic Resource Discovery, is a specification published June 17, 2026 by Google, Microsoft, GoDaddy, Salesforce, and others.
It defines a small manifest file, ai-catalog.json, that a business publishes on its own domain — similar in spirit to robots.txt — plus a registry API that AI agents use to discover and verify a business’s services, pricing, and availability.
Do small local businesses need to worry about agent-to-agent commerce yet?
Not to panic about, but yes to prepare for.
The infrastructure — A2A, ARD, Gemini Spark — is live and adopted by major platforms today.
A small trades or home-service business does not need to build any of this itself, but it does need the underlying structured business data in place, because that data is what an agent will read once it starts asking on a customer’s behalf.
What should a local business do to prepare for AI agents booking jobs?
Get the foundational structured data right first: accurate, detailed service and pricing information, a complete Google Business Profile, and consistent business facts published across the web.
Agents can only transact with what they can verify, and verifiable structured data is the prerequisite for everything the newer protocols build on top of.
About Stellaris Ridge
Stellaris Ridge builds AI automation for local trades and service businesses.
Our AI voice agent, VOX, answers every call 24/7 and books the job before voicemail can lose it — backed by missed-call recovery and follow-up that runs in the background.
- Jarod Treppish, co-founder — the face of the company and the person you’ll actually talk to.
- We work with owner-operated shops — local trades, 1-15 employees.
- Built and run by a team that ships. When you win, we win.
→ See what Stellaris Ridge can do for your shop
Related reading
- Google Gemini for Small Business — the practical starting point for using Google’s AI tools day to day.
- GEO vs SEO for Local Trades — how citation-based search changes what you publish, and why both still matter.
- How AI Search Decides Which Local Businesses to Recommend — the signal set behind every AI recommendation, human or agent-driven.
Also On Our Network
- 🌐 Your Website Isn’t the Front Door Anymore — the zero-click search numbers driving this whole shift.
- 🌐 The Knowledge Catalog: The Asset That Gets You Found — the hands-on build guide for the structured data agents actually read.
📞 Never miss a lead call again — see VOX in action 👉
🛠️ You handle the pipes. We handle the AI. Talk to Stellaris Ridge 👉
AI agents booking home service jobs is not a future headline anymore — it is live infrastructure today, and the businesses that show up in that flow are the ones whose data is ready for it now.