A Handyman Who Uses AI to Diagnose Jobs, Find Rare Parts and Build His Own Tools: Ray Duke
Ray Duke runs a handyman business near Tucson, Arizona, working for property managers. He uses AI to work out thermostat wiring from photos, reset a luxury fridge drawer instead of replacing it, track down discontinued parts, soften emails to unhappy clients and build his own pricing tools without knowing how to code. He is also clear about the limits: AI can show him the price range in his market, but the price is his decision.
A handyman's day is full of problems nobody warned them about: a thermostat with a tangle of unlabeled wires, a part that stopped being made decades ago, an unhappy client who needs a careful reply. For someone working alone, every hour spent researching is an hour not spent on paid work.
Ray Duke, a handyman near Tucson, Arizona, uses AI for exactly those moments. He describes it as a tool like any other, one that helps him finish jobs he might otherwise have had to turn down.
This article is for tradespeople and anyone who runs a hands-on service business. We look at how he uses AI on the job and in the office, what he refuses to hand over to it, and then, in a separate section, at what his experience still has to do.
Who he is
Ray Duke runs a handyman business near Tucson whose clients are property managers. He also runs Bulletproof Handyman, a YouTube channel, newsletter and website where he teaches other tradespeople how to run their businesses and shares free tools he has built. In an August 2026 issue of his newsletter, he wrote that he is now the only person doing maintenance work at his company.
The details of how he uses AI come from a December 2025 webinar he gave for Jobber, the field service software company he uses.
How he uses AI on the job
- Diagnosing problems from photos. On one job, he arrived to replace a thermostat and found only a bundle of wires coming out of the wall. He photographed the wires, the new thermostat and its model number, and the air conditioning unit on the roof, and gave them to ChatGPT. It told him where each wire should go, and it was right. He says that turned what could have been an eight-hour job into a one-hour one.
- Saving the client a replacement. At an expensive house, a built-in fridge drawer had stopped working and was due to be replaced for about $800 plus an appliance technician's labour. He gave AI the fridge's model number, and it told him which buttons to press to reset the drawer's software. The drawer worked again.
- Finding rare parts. For a 1950s house, he needed a valve stem from a manufacturer that had gone out of business. Local stores did not recognise it. AI identified the part and found what he believes were the last two available for sale online in the United States, sparing him from opening up the wall. On another job, he set an AI agent to search the web for unusually narrow mini blinds; after about half a day of occasional check-ins, it found the one site that sold them.
- Showing clients the result. He has tested giving AI a photo of a client's kitchen and asking it to show the room with different cabinets, which he says is good enough for a client to decide whether they like a direction.
How he uses AI in the office
- Difficult emails. When a client is unhappy, he gives AI the whole email thread, explains the situation and the tone he wants, lets it draft a reply, then edits it himself so that he is the last one to touch it and it still sounds like him.
- Quotes tailored to the client. He suggests telling AI whether a client cares most about quality or price, and asking it to write the quote with that in mind.
- Marketing. He thinks AI writes advertising copy better than almost anything else, and suggests describing your business to it to get seasonal service ideas.
- Comparing tools. For important tasks, he gives the same prompt to ChatGPT, Grok and Gemini and compares the results.
- Building his own software. Although he says he cannot code, he builds software for other tradespeople by describing it to AI in plain English. His website offers free tools such as an hourly rate calculator.
What he does not hand over to AI
- Setting his prices. He tells tradespeople that AI cannot tell them what to charge. It can show the range of prices in a local market, but where a business sits in that range depends on its quality and clients, and that decision is the owner's.
- Legal and tax decisions. He finds AI useful for drafting terms and conditions and thinking through tax planning, but says anything important should still be checked by a lawyer or another professional.
- Learning the trade. He admits that when AI guided him through fixing a video doorbell, he still does not know what he actually did. He warns that tradespeople should keep learning rather than handing their thinking over to AI.
What trade experience gave him that AI could not
This section is our analysis, not a claim made by Duke.
- Knowing what to photograph. AI could read the wiring only because he knew which parts of the system mattered and how to describe the problem.
- Judging whether the answer is safe. Electrical and plumbing work carries real risk. Deciding whether an AI suggestion makes sense on site depends on hands-on knowledge.
- Knowing the client. Choosing whether a quote should stress quality or value, or how to calm an unhappy client, comes from years of working with property managers.
- Owning the final word. His habit of editing every AI draft himself keeps his voice, and his responsibility, in every message.
What you can copy
These are our suggestions, drawn from his story.
- Give AI the full picture. Photos, model numbers and a clear description of the symptoms produce far better answers than a vague question.
- Use AI for research you would never have time for. Tracking down a discontinued part is exactly the kind of task it can grind through.
- Let AI draft, then rewrite it yourself. Especially for sensitive messages, make sure you are the last one to edit.
- Ask for a range, not a price. Use AI to understand your market, then set prices yourself.
- Keep learning the trade. When AI solves a problem for you, take time to understand what was actually fixed.
Notes on the sources
The details of how Duke uses AI come from a webinar he gave for Jobber, the field service software company he uses, recorded on December 2, 2025; the page includes a full transcript. Jobber hosted the event to promote its software, and Duke speaks warmly of Jobber in it. The information about his current business comes from his own website and newsletter. In a December 2024 issue of the newsletter, he wrote that he planned to move from subcontractors to employees in 2025, starting by hiring his son; later issues we reviewed do not say whether that happened, and in August 2026 he wrote that he is the only person doing maintenance work at his company. The webinar does not mention employees. Neither source gives revenue figures.
Sources
- 10 Ways to Use AI and Save Hours of Work, Jobber (webinar with Ray Duke, recorded December 2, 2025)
- About, Bulletproof Handyman
- Bulletproof Handyman Newsletter, August 2026 (Volume 18)
- Bulletproof Handyman Newsletter, December 2024 (Volume 12)
Related articles
Showing Clients the Room Before It Exists: Two Independent Interior Designers and AI Visuals
Interior designers sell spaces that do not exist yet, and photorealistic renderings have long been expensive. A sole-practice interior designer with nearly 30 years of experience uses an AI tool to take his own renders to a quality he could not reach before. A Seattle designer uses ChatGPT to show clients ideas they would never pay a renderer for. Here is what each one hands to AI, and what stays with the designer.
Notes Done Before the Patient Leaves: A Solo Physiotherapist and a Solo Psychologist on AI Scribes
For clinicians in solo practice, documentation is unpaid work that often spills into evenings. A physiotherapist in Queensland, Australia, now completes most of her notes while she treats, and says she ended her after-hours paperwork within two weeks. A clinical psychologist in Singapore saves up to 45 minutes on each new intake note and finds she can stay more present with clients. Here is what each one hands to AI, and what stays with the clinician.
Two Jewellers Who Make Everything by Hand and Let AI Help with the Words: Leto Lama and Sarah Kobak
Independent jewellers sell objects made by hand, but the business around them runs on words: captions, product descriptions, stories and messages to clients. A London goldsmith who makes every piece herself uses ChatGPT for captions and content ideas and AI for bespoke renders and gold calculations, and says it has helped her make far more jewellery. A Dutch goldsmith writes everything herself and uses AI only to rethink how she says it. Here is what each one hands to AI, and what stays in their hands.
One Planner Uses AI to Save Time on Event Paperwork, the Other Uses It Sparingly: Two Independent Event Planners
Event planners juggle rooming lists, budgets, menus, contracts and minute-by-minute schedules. Two independent planners interviewed by the events association MPI take very different approaches to AI. One uses it to save time on rooming lists, budgets, manuals and schedules. The other uses it sparingly, never uploads contracts or attendee data, and worries about its environmental cost. Here is what each one hands to AI, and what stays with the planner.