The Hardest Part of Running a Small Business in the Trades
The AI revolution has triggered a massive shift in daily life for knowledge workers. Developers, writers, analysts, and designers have seen their output transform dramatically over the past year. But that revolution is still working its way into the industries that rely less on sitting at a desk. The trades are one example: skilled, independent contractors running their own businesses have enormous amounts to gain from AI, but the tools built so far weren’t built with them specifically in mind.
At Mozilla.ai, we think about trust, transparency, and user agency as foundational to what good AI looks like. It means building for the people who’ve been left out of the current wave, not just the people already in front of a screen. That’s what led us to Clawbolt.
The Problem
There’s a story that plays out constantly in the trades. Someone spends years working for a larger company, gets great at their craft, and eventually makes the leap: be your own boss, do great work, and experience the benefits from the effort you put in.
What they quickly discover is that running a business is a whole lot more than being good at your craft. Wrapped around the work is a mountain of administration:
- Visiting job sites to give quotes and estimates
- Researching the cost of materials
- Planning and managing schedules
- Hiring and coordinating day laborers or subcontractors
- Sending and tracking invoices
- Processing payments
- Managing business profiles, reviews, and social media
Every hour spent at a keyboard chasing invoices or updating a business profile is time not spent on the jobs that are generating the revenue. This is why so many small businesses in the trades struggle. The skill is there, but the bandwidth for everything else often isn’t. And when the business is going well, the “reward” is frequently an evening in front of a laptop catching up on paperwork instead of time with family.
Why AI Agents, and Why Now?
Most people are familiar with AI assistants in the ChatGPT mold: you ask a question and you get an answer. It’s useful, but it puts the burden on the user to know what to ask and when to ask it.
That changed this fall with the emergence of OpenClaw, an open-source project that became the highest-starred Github repository of all time. OpenClaw introduced a framework for AI that operates proactively in the background, taking initiative, surfacing things the user didn’t know they needed to handle, and acting on their behalf without waiting to be prompted.
The catch (and it’s a big catch) is that OpenClaw is hard to set up, and misconfiguration has massive security implications. It’s a powerful foundation, but it’s not something most people can just pick up and use safely.
Introducing Clawbolt.ai
Clawbolt is an idea we’ve started working with at Mozilla.ai: a narrow, purpose-built AI assistant for contractors and small trade business owners. It’s not trying to be a general-purpose tool. It’s designed around the specific, repeatable needs of someone running a small trade business without a back-office team to support them.
A few of the guiding principles of Clawbolt:
- It meets users where they already are. The interface is a messaging app they already use, whether that’s Telegram, WhatsApp, or iMessage. No new software to learn, no browser tabs to manage. The user experience is designed from the ground up to work just like you’re messaging a friend: scheduling reminders, approving data access, updating configuration. All designed to happen smoothly over messaging apps.
- It connects to the tools they already use. Clawbolt integrates with accounting software like QuickBooks and with calendar apps like Google Calendar to handle scheduling and finances without requiring the user to leave their conversation thread.
- It’s proactive, not passive. Rather than waiting to be asked, Clawbolt learns where a particular user tends to fall behind and gets ahead of it. That might mean following up on an unpaid invoice, flagging that material costs have changed on an active bid, or reminding someone to schedule a follow-up call.
- It’s built on open-source foundations with security as a priority. Mozilla.ai’s commitment to transparency means Clawbolt has an open source core, and we’re taking our time with curating integrations to ensure that security isn’t an afterthought.
We’re also working on a hosted option for people who want to get started without any technical setup. Self-hosting shouldn’t be a prerequisite!
Our shining star: a contractor should be able to finish a long day of work, go home, and not have to spend hours on a laptop to get paid for work they already did.
Get Involved
Clawbolt is still in early development, and that’s an intentional decision: the earlier we hear from people working in the trades, the more that input can shape how we build.
If you work in the trades, manage a small trade business, or know someone who does and any of this resonates, we want to hear from you. You can fill out this quick form or reach us at [email protected]. If you’re a software developer and want to dig into the project, contribute, or give it a star, the codebase is public on github.com/mozilla-ai/clawbolt.
Mozilla.ai is a public benefit startup and wholly-owned subsidiary of the Mozilla Foundation, operating with its own independent team. Our work focuses on AI technologies built around agency, access, and transparency. We share the Mozilla name and values, but we’re a separate organization from Firefox, Thunderbird, and other Mozilla products.
Curious about the Mozilla family? Mozilla Foundation · Mozilla Corporation · Mozilla Ventures · Mozilla Data Collective · Firefox · Thunderbird
This article first appeared on Read More


