KnowHow Is Building the Field Manual That Tradespeople Actually Want

The Macro: A Million Skilled Workers Are About to Retire

The skilled trades are facing a workforce crisis that’s been building for two decades and is now hitting full force. The average age of a licensed electrician in the US is 55. Plumbers, HVAC techs, welders, and carpenters are all trending the same direction. The Bureau of Labor Statistics projects that construction alone will need to attract roughly 500,000 new workers per year just to keep pace with retirements and demand growth.

This isn’t a new observation. What is new is how little progress has been made on the knowledge transfer problem. When a master electrician with 30 years of experience retires, that knowledge walks out the door. Sure, there are training programs and apprenticeships. But the practical, job-specific knowledge, the kind that says “when you’re doing a panel upgrade in a 1970s ranch house, watch for aluminum wiring and double-tapped breakers,” that stuff lives in people’s heads, not in manuals.

The software industry hasn’t done much for tradespeople. Procore dominates project management for large commercial jobs. Buildertrend and CoConstruct serve residential contractors. ServiceTitan handles field service management for HVAC and plumbing companies. But these are all workflow tools. They track jobs, invoices, schedules, and customers. None of them solve the fundamental problem of getting knowledge out of experienced workers’ heads and into the hands of the people who need it at 2 PM on a Tuesday when they’re standing in front of a junction box they’ve never seen before.

The reason this hasn’t been solved is that capturing and organizing trade knowledge is genuinely hard. Every job is different. Code requirements vary by jurisdiction. Best practices change based on the specific equipment, building age, and conditions on site. A static manual can’t account for that variability. But an AI that can generate contextual, step-by-step instructions based on the specific situation? That’s a different proposition entirely.

The Micro: Guides That Meet Workers Where They Are

KnowHow is an AI assistant built specifically for the trades. The core product does two things: it uses AI to generate step-by-step guides that tradespeople can reference while they’re actually on the job, and it helps general contractors and project managers keep track of their teams and projects.

The company is based in San Francisco and came through Y Combinator. The website was unresponsive during my research, which could mean they’re in the middle of a redesign, dealing with traffic spikes, or just running on infrastructure that wasn’t built for high availability yet. None of those are disqualifying at this stage.

What I can piece together from the product description is a two-sided value proposition. For the worker in the field, KnowHow provides AI-generated instructions that are specific enough to be useful. This isn’t a chatbot giving you generic advice. It’s supposed to produce actual step-by-step guides that account for the specific task, tools, and conditions. For the GC or project manager, it’s a management layer that helps coordinate teams and ensure work is being done correctly.

The AI angle here is interesting because trade work has a lot of the characteristics that make AI generation useful: high variability (every job is different), a large body of underlying knowledge (building codes, manufacturer specs, best practices), and a delivery format that benefits from customization (steps tailored to the exact situation). A plumber installing a tankless water heater in a new build follows a different set of steps than one retrofitting the same unit into a 1960s colonial with galvanized pipes and low water pressure.

The competitive set is thin. SkillCat offers training and certification for HVAC and electrical workers. CompanyCam helps contractors document job sites with photos. But nobody else is generating contextual work instructions using AI for tradespeople. That’s either a sign of a massive untapped opportunity or a sign that the market hasn’t shown enough willingness to pay for this type of solution. I think it’s the former, mostly because the workforce crisis is creating urgency that didn’t exist five years ago.

The biggest open question is adoption. Tradespeople are pragmatic. If the guides are accurate and save time, they’ll use them. If the guides are generic or wrong, they’ll delete the app in about 30 seconds. There’s zero tolerance for software that wastes time when you’re on a job site with a deadline.

The Verdict

KnowHow is attacking a real problem at the right time. The labor shortage in skilled trades is not theoretical. It’s happening right now, and every contractor I’ve talked to in the last year mentions it as their top concern.

At 30 days, I’d want to see the quality of the generated guides. Pull five random guides from different trades and show them to experienced workers. If those workers say “yeah, this is right, and it would help a newer guy,” that’s signal.

At 60 days, the question is whether GCs and project managers adopt the management features. Getting the field workers to use the guides is step one. Getting management to rely on the platform for coordination is step two, and that’s where the revenue likely scales.

At 90 days, I’d want to understand the data flywheel. Every guide that gets used in the field is an opportunity for feedback. Does the AI learn which steps are confusing, which get skipped, and which cause callbacks? If so, the guides get better over time, and that compounds into something very hard to replicate.

The trades need this. The question is execution, and that comes down to whether the guides are good enough to earn trust from people who don’t give trust easily.