The Macro: The Trades Are Booming and the Back Office Is on Fire
There is a labor shortage story that gets told constantly in tech circles, and it goes something like this: AI is coming for knowledge worker jobs, and we should all be worried. But there is a different labor shortage that is arguably more acute and gets almost no attention from the startup world. Skilled tradespeople are aging out of the workforce faster than they are being replaced, and the businesses they run are drowning in administrative work that has nothing to do with their actual skills.
The numbers are straightforward. The Bureau of Labor Statistics projects a shortage of over 500,000 skilled tradespeople in the US by 2028. The average age of a plumber in America is 49. Electricians, HVAC technicians, and general contractors are in a similar demographic crunch. These are people who are excellent at their craft and frequently terrible at running a business, not because they are not smart but because responding to leads within five minutes, generating accurate estimates, coordinating material procurement, and keeping up with invoicing are full-time jobs in themselves.
Most trades businesses solve this problem one of two ways. They hire an office manager (expensive, hard to find, single point of failure) or the owner does it themselves (unsustainable, leads to burnout). There is a reason the failure rate for trades businesses in the first five years is over 50%. It is rarely because the work is bad. It is because the business side eats them alive.
This is the kind of problem AI should be solving, and almost nobody in the startup ecosystem has been paying attention to it. The VC money goes to developer tools, fintech, and enterprise SaaS. Meanwhile, there are millions of trades businesses running on text messages, paper estimates, and a spreadsheet they set up in 2019 and have not updated since.
Ressl AI is going after this market with a simple pitch: AI employees that handle the back-office work so tradespeople can focus on the trade.
The Micro: Former Google Scholar Meets IIT Bombay in the Service Van
Ressl AI came through Y Combinator’s Winter 2026 batch, founded by Arushi Gandhi and Abhishek E. Gandhi is a former Google Scholar and ex-Microsoft. Abhishek is an IIT Bombay graduate. On paper, this is a team with serious technical credentials building for an industry that most of their Stanford and MIT peers would overlook entirely. That disconnect is actually an advantage. The incumbents in this space, ServiceTitan, Housecall Pro, Jobber, are all workflow tools. They digitize the process. Ressl is trying to automate the process.
The product positions itself around the concept of “AI employees.” Not copilots. Not assistants. Employees. The language is deliberate. A copilot suggests you are still doing the work with some help. An AI employee suggests the work gets done whether you are looking at your phone or not.
From the YC listing, the agents cover lead response, estimating, procurement, and general administrative tasks. Let me break down why each of these matters.
Lead response is probably the highest-leverage problem on the list. In home services, the first business to respond to a lead wins the job something like 78% of the time. Most contractors respond in hours, not minutes. An AI agent that can engage with a lead immediately, ask qualifying questions, and schedule an estimate could meaningfully change the conversion math for a small business.
Estimating is where the domain knowledge gets deep. A roofing estimate is not the same as a plumbing estimate. Material costs vary by region. Labor rates vary by market. An AI that can generate accurate estimates needs access to real pricing data and trade-specific logic, which is hard to build but enormously valuable if you get it right.
Procurement is the sleeper category. Most contractors order materials by calling their supplier rep or driving to the supply house. An AI that can auto-generate material lists from estimates and place orders with preferred vendors saves hours per week and reduces waste from ordering errors.
The competitive picture includes ServiceTitan (the 800-pound gorilla, valued at over $9 billion), Housecall Pro, Jobber, and FieldPulse. But these are all SaaS platforms that require the business owner to log in and use them. Ressl is positioning as something that works autonomously. That is a fundamentally different value proposition for an owner-operator who is on a roof all day and cannot check a dashboard.
The Verdict
I think this is one of the most interesting applications of AI agents I have seen, precisely because the target customer is so far from the typical startup buyer. A plumbing contractor in Phoenix does not care about AI. They care about whether their phone rings and whether the estimate goes out on time. If Ressl can deliver on that promise in language the customer understands, the market is enormous and underserved.
The challenge is go-to-market. Reaching trades businesses is notoriously hard. They do not read TechCrunch. They do not attend SaaStr. They learn about new tools from their suppliers, their trade associations, and other contractors they trust. Ressl will need channel partnerships and boots-on-the-ground sales tactics that look nothing like a typical SaaS launch.
I would also watch the accuracy of the estimating agent closely. If it generates an estimate that is 20% low, the contractor eats the difference. If it is 20% high, they lose the job. The tolerance for error in this industry is very thin, and the consequences of getting it wrong are immediate and tangible in a way that is different from most software markets.
The founding team has the technical chops. The market has the pain. The question is whether they can build trust with an audience that is justifiably skeptical of technology that promises to do their job for them. If they can, this is a billion-dollar company hiding in plain sight.