How I Automated My Property Services Business with OpenClaw

I'm Not a Tech Person

Let me start there. I'm a Massachusetts-licensed tradesman (CS#115374). I know how to fix things, talk to clients, and manage schedules. I don't have a computer science degree. But about a year ago, I hit a wall with my property services business, Smith & Company.

I was doing everything myself—answering emails at midnight, tracking invoices in spreadsheets, manually following up with clients, scheduling jobs, monitoring what was actually happening in the business. There was no system. Just me, overwhelmed, trying to keep plates spinning.

The moment I realized things had to change was when I missed a payment reminder for a client, lost a potential job opportunity, and spent three hours that day just digging through emails looking for a single invoice number.

The Turning Point

Around the same time, I started hearing about AI agents and the OpenClaw framework—an open-source system designed to let AI systems actually do work, not just chat. It sounded like science fiction. But I kept seeing people in my network talking about what it could do.

I spent a few weeks researching, testing, breaking things (a lot), and eventually building a system that now runs the back office of my business. Not a chatbot that I talk to. An actual working system that monitors emails, sends daily operation briefs, tracks invoices, schedules follow-ups, and generates content.

What I Actually Built (Day-to-Day)

Here's the honest version—what my system actually does in real operation:

• Email monitoring: It reads incoming client emails and sorts them by priority. If something needs immediate action, I get alerted. Otherwise, it logs it for the brief.
• Daily operations brief: Every morning, I get a concise summary—open jobs, pending invoices, client messages that need responses, scheduling conflicts, anything that needs my attention today.
• Jobber integration: This was custom. I built an MCP (Model Context Protocol) connector to my Jobber system so the AI actually has real-time data about jobs, scheduling, and customer history.
• Invoice tracking: It monitors which invoices are outstanding, generates gentle reminders, and flags ones that are aging beyond 30 days.
• Client follow-ups: After a job completion, it automatically generates personalized follow-up messages. Not templated. Actually personalized based on the job details.
• Scheduling automation: It coordinates with my calendar, looks at travel time between jobs, and suggests optimal scheduling. Sometimes I accept, sometimes I adjust.
• Website and content: Blog posts, service updates, seasonal content—it helps generate the copy, and I refine it.

The Stack: OpenClaw + NemoClaw (Explained Simply)

I'll keep this non-technical because honestly, I'm still learning this stuff too.

The point of this architecture isn't to be complicated—it's to be reliable. If one thing breaks, the other keeps running. If one server gets slow, the load distributes.

What's Actually Working

I want to be honest here. Not everything is perfect.

The email monitoring is solid. It catches maybe 95% of what I'd catch manually, and it doesn't miss urgent stuff. That alone has saved me hours per week and prevented me from missing follow-ups.

The daily brief is genuinely useful. I read it with my coffee every morning, and it changes how I prioritize my day. Instead of reacting to whatever lands in my inbox first, I'm working off actual business data.

Jobber integration works. Real-time job data means nothing gets stuck in cracks. Scheduling automation caught a double-booking error last month that would've been a mess with a client.

Invoice tracking is maybe 90% there. The reminders work, but sometimes clients pay between when the reminder is sent and when I see it—so I have to manually mark it complete. Small issue, but honest one.

What's Still In Progress

The content generation is rough. It needs my editing. Sometimes it doesn't capture the voice I want or misses nuance about a specific service. I usually rewrite the first draft. But it cuts my writing time from 2 hours to 30 minutes, so it's still valuable.

Client follow-ups sometimes feel a little too polished coming from an AI system. I'm working on making them feel more like they're actually from me. It's a balance between automation and authenticity.

Scheduling optimization is good at seeing the puzzle, but sometimes it doesn't account for client preferences or job complexity the way I would. So I always review suggestions before confirming.

Why I'm Building Logos HQ

I started getting emails. Contractors in other states, service business owners, tradespeople like me asking, "How did you do that? Can you help me build this?"

At first, I thought I'd just give advice. But then I realized—a lot of businesses like mine never even consider this because it seems too technical. It seems like something only tech companies do. But it's not. It's actually for businesses like mine. For people like me.

So I'm launching Logos HQ. It's a consulting business to help other trades and service businesses build their own versions of what I built. I'm not trying to sell you software. I'm trying to share what I learned so you don't have to figure it out alone.

What I'd Tell You If We Were Having Coffee

If you're running a service business and you're exhausted from doing everything yourself—automation is not something you do later when you're bigger. It's something you should do now because it makes you bigger.

It doesn't have to be perfect. My system isn't. But it works. It actually works.

And you don't need to know code. I didn't. I'm learning as I build, and I'm still making mistakes.

Next Steps

I'm sharing this not as the end of the story, but as the middle. There's a lot more to come—better scheduling, predictive insights, customer portal integration, probably things I haven't thought of yet.

But the core insight is already here: a tradesman in Massachusetts built an AI system to run his back office, and it actually works. If that sounds interesting, let's talk.