TradeTalent

Overview

Construction has a serious math problem. Roughly five skilled workers retire for every new one who enters the trade, and the industry has been short hundreds of thousands of people for years. Projects slip or get turned down, not because the work isn’t there, but because the crew isn’t.

The companies feeling this hardest are small and mid-sized construction firms. They range from 50 to a few hundred employees, bring in $8 million or more in revenue, and often have an HR function of just one or two people. In the segment we targeted, that was roughly 64,000 companies in the US.

Those teams were trying to solve a hiring crisis with personal networks that had run dry and software built for an entirely different kind of candidate.

TradeTalent is an AI recruiting platform purpose-built for construction and the skilled trades. It turns any physical location into a sourcing channel via QR codes and text-to-apply, scores incoming resumes against the job's actual requirements, and screens candidates through AI text conversations rather than phone calls. Strong candidates reach a hiring manager within hours, not weeks.

I co-founded TradeTalent and led it from problem identification and customer discovery through the first product build and pre-seed investment.

The Challenge

The prolonged labor shortage was easy to spot. The interesting part was its impact on the person responsible for hiring and why current solutions were failing both companies and candidates.

We interviewed dozens of construction HR professionals and spent days on-site watching some work. We found HR teams that were understaffed, time-constrained, and left to do it all. One professional described herself as doing “everything except swinging the hammer,” which meant HR, accounting, scheduling, and management all landed on her desk. Recruiting competed with every one of those tasks, and always lost.

These teams either used horizontal HR platforms built for white-collar hiring or ran everything manually. Referrals and personal networks did the heavy lifting, and that worked for years, until the labor shortage made sure it didn’t. The most common thing we heard was that a network that had always produced people had simply dried up.

Simply driving more applicants wasn’t the answer either. We learned professionals often wasted more time hiring after a hiring post started working because volume without soft and hard-skill filtering is often worse than no candidates at all. We found candidates regularly cleared an entire process only to be unqualified when they connected with a foreman or, worse, arrived on the jobsite ready to work.

Underneath it all, hiring for skilled trades is just different. Skilled tradespeople often have incomplete resumes or no resume at all, small digital footprints, and English language barriers. At the same time, traditional recruiting software assumes a clean resume, an online presence and application, and an HR team large enough to run a pipeline. It is built around artifacts these candidates don’t have and processes that don’t fit, which is why horizontal tools kept getting hired for the job and then abandoned.

The Solution

TradeTalent was built around a simple bet: in construction, HR software should fit the industry's and candidates’ unique needs and do the work rather than organize it.

Traditional applicant tracking is a system of record. It helps you track candidates you already have through stages you already run. That is the wrong shape for a two-person HR team that's run out of sourcing options and is drowning in work.

TradeTalent is an AI recruiting platform purpose-built for the construction industry. It sources, screens, and qualifies candidates to help busy HR teams close their skilled trades hiring gap by finding qualified candidates faster, with less work.

Sourcing meets candidates where they are. A QR code on a jobsite sign, a truck door, or a supply house counter opens an AI text conversation instead of an online application. Any physical location becomes a recruiting channel, and the biggest drop-off point disappears for people without a resume.

Everything after sourcing compresses a sequence that used to run on manual effort. Resumes are read and scored against the actual job requirements as they land, with the reasoning visible, so recruiters have a credible ranked shortlist. Chat-driven screening replaces the endless pre-screening phone calls with a text conversation that shares the requirements and collects the details a recruiter would otherwise spend a week chasing down via voicemail. Auto scheduling syncs the team’s calendars and coordinates the interview by text once a candidate clears the bar.

The design constraint we held onto came directly from the research. People told us they were willing to give up some control to get their time back, but only if they stayed confident that the qualities they cared about were being measured. So the user defines the criteria, the scoring is explained, and no one gets rejected without human approval.

The Approach

Like most ideas, we found this one while exploring a completely different problem: materials sourcing and financing for sub-contractors. We tested a concept in that space and kept hearing that labor was the bigger pain, so we quickly pivoted.

We focused on demand first, and ran multiple parallel workstreams. Deep desktop and competitive research, exploratory and JTBD interviews, a day-in-the-life study, and a review of venture investment activity in the space.

We gathered information to define the problem space, segmenting the labor space to understand the nuances of what made up construction labor. We explored the solution and competitive landscape, sorting the field into horizontal recruitment tools, applicant tracking solutions, recruiting-as-a-service firms, and emerging workforce-as-a-service platforms, and compared them across business attributes and the full hiring workflow.

The exploratory interviews and onsite study helped us map the hiring journey across finding, vetting, and retaining; we connected each problem people described to the activity, and had industry experts review the map with us to identify what we might have missed. We found that finding and initially vetting candidates were the areas teams struggled most.

The jobs-to-be-done sprint sharpened our understanding and grounded it in what people actually did, not just what they said they did. Three jobs surfaced, and we built exclusively for the first, helping HR teams quickly find and qualify the candidates they need. We defined what basic quality meant for the initial target customer and baked it into early user stories and a short feature set, sorted by table stakes, performance, and delighters. This work set up a loose but testable product direction we could iterate on quickly.

To de-risk the venture, we ran multiple parallel experiments using design, functional, and resume-scoring prototypes; we spoke with decision-makers and candidates; ran paid ads and cold email outreach to test value props and get early acquisition signals; and we built design partner relationships that fueled our early customer pipeline, helped us refine our early ICP and product roadmap, and built industry credibility.

Decisions

Decisions we made with high levels of ambiguity and risk that helped us learn and move forward

Vertical, not horizontal.

Over the years, labor platforms have been evolving from shallow horizontal marketplaces focused on matching to deeper job platforms where discovery, vetting, and other tools add value. We saw the opportunity to build a construction-first vertical product that removes the unique friction points facing the construction industry and ship something differentiated rather than something broad and forgettable.

Do the work, don’t just record it.

We decided the right move was to automate the manual labor in front of the applicant tracking system, not just build a better applicant tracking system for construction. That decision cost us initial clarity, with users asking whether we could replace their ATS. Still, it solved a real problem for construction firms and saved us from competing on feature count against traditional horizontal ATS solutions with a decade-long head start.

Source where candidates actually are.

Tradespeople live on their phones and rarely have polished resumes, while traditional hiring tools still assumed desktops and polished resumes. We bet on QR codes and AI-driven text conversations instead of job boards and human screening calls.

Ship a wedge first.

Waiting for a “finished product” to learn through and sell would have been the most expensive mistake an early venture can make. Leading with a standalone resume-scoring tool let us generate leads, refine and prove the resume-scoring model, and learn from real usage months before the full product existed.

Results

Raised a $3M seed round.