Jeff Graham is building software to automate the counting no one should be doing by hand — and a staffing company that puts Maritime tradespeople to work across Canada. The two businesses look unrelated until he explains the thread.
On paper, Construction AI and Blueforce Logistics have nothing in common. One is a deep-tech software play run out of Nova Scotia, converting PDF drawing packages into machine-readable data for estimators and engineers. The other is a skilled-trades placement company that recruits welders, CNC machinists, and conveyor assemblers from the Maritimes and flies them to projects in British Columbia, Ontario, and beyond. Jeff Graham, a communications professional turned construction-tech founder, runs both.
The thread connecting them is straightforward: skilled people are doing work that doesn't need them, and someone should fix that. A senior estimator counting doors and window openings on a PDF is doing data entry. A qualified welder in Pictou County sitting idle while a plant in BC can't fill a contract is a logistics problem. Graham's answer to the first is software. His answer to the second is a phone call and a plane ticket.
He sat down with host Daniel Arsenault on Episode 8 of the Atlantic Construction Podcast. Here is what a contractor should take from the conversation.
Automate the counting, not the judgment
Construction AI's core product starts with a problem every estimator knows: the bulk of a take-off is mechanical. Count the doors. Count the windows. Extract the earthworks elevations. Record the MasterFormat codes. None of that requires a senior estimator's judgment — but it consumes an enormous portion of their working day.
Graham's software reads PDF drawing packages and pulls out quantities, geometry, object counts, and CAD data automatically. For earthworks, it extracts point-cloud cross-section data and exports it to DXF or CSV formats a grader operator can use. The pitch is not that the software replaces the estimator; it's that the estimator should be pricing the work and writing the strategy, not counting what's already drawn.
His framing is blunt — "don't let them spend their time counting stuff that's ridiculous" — and it holds for any repetitive, rule-bound task an experienced person is currently stuck doing by hand.
The software uses supervised machine learning, not rule-based automation. The distinction matters operationally. A rule-based system processes inputs the same way every time; a supervised model learns from examples, which means it can handle the variation in real drawing packages. It also means it can fail in ways a rule-based system wouldn't. Graham is direct about the maintenance overhead this creates: "when it fails at its task a data scientist looks at it and goes okay why did it fail" The humans don't disappear from the loop; they move from counting to quality control.
The co-founder handling that technical layer is Dr. Jason Heard, a computer scientist at Mount Royal University in Calgary with a background in multi-agent systems and genetic programming. Construction AI also carries strategic investors from the modular and residential construction sectors — people with production knowledge Graham didn't start with. Their value, as he describes it: "they're kind of able to educate me on some of the knowledge gaps that we have".
Get the distribution question right before building the sales channel
Construction AI is not going to market by hiring a direct sales team and cold-calling estimating departments. The commercialization plan is to slot into the workflow of companies that have already won that relationship — specifically platform vendors like Trimble and Hexagon AB, which provide positioning and construction-software infrastructure to major contractors and project owners worldwide.
The logic is practical: large platform vendors have active customers who already rely on them for estimating workflows. An AI module that improves take-off speed is a natural add-on for those platforms to offer. Building a distribution channel from scratch takes years and capital; building a component another platform distributes is faster and cheaper.
The tradeoff is dependence on a partner's roadmap and deal timelines. But for a deep-tech startup with a technical moat and limited go-to-market budget, it is a credible path to the market.
Nova Scotia is a workable base for deep tech
Graham makes a concrete case for building a tech company in Atlantic Canada. The province has a cluster of funding and support programs: Innovacorp (now Invest Nova Scotia), ECOA, CBDC, NRC IRAP, and the SR&ED federal tax credit for R&D. He points to "a lot of really good government support through like anova core ecoa uh there's the cbdc". The universities — Dalhousie, Saint Mary's, StFX — produce data scientists and AI talent. The culture is welcoming to entrepreneurs rather than skeptical.
The operating cost advantage is real. A software team in Halifax costs less than the same team in Toronto or Vancouver. For a company that needs to run long enough to prove a technically hard product, that runway extension is not a minor consideration.
When you win more than you bid for, you need flex labour
Blueforce Logistics started from a single conversation. An investor in BC couldn't staff a project fast enough; the local labour market didn't have the specific tradespeople he needed on the timeline he needed them. Graham built the company to solve exactly that problem at scale.
The service model: Blueforce recruits tradespeople from across Canada, handles all travel logistics, and places them on contract projects for client companies. Workers go out on eight-week rotations. They are not "out of pocket a dime right so it's like they just hop they basically hop on a plane" — flights, ground transport, and accommodation are all covered. The worker shows up ready to work.
From the client side, the use case that shows up most often is surge hiring. A contractor bids conservatively and wins more than expected: "we bid three jobs we thought we would get one but we got all three so now we need to ramp up". Permanent headcount added for a surge creates an overhang when the surge ends. A flexible placement solves the problem without locking in the overhead.
For hard-to-fill trades — CWB-certified welders, CNC machinists, specific electrical certifications — a national recruiter has a structural advantage over a local temp agency. A local firm draws from within commuting distance. Blueforce will "scour every square inch of canada" to find the right person. The Maritime provinces are a particularly productive source: experienced tradespeople, strong work ethic, and a candidate pool that the national market hasn't fully tapped.
Why Maritimers answer the phone on Facebook
Blueforce recruits primarily through Indeed and Facebook. The Facebook choice is not default — it is deliberate and grounded in platform penetration data. Facebook's reach in Atlantic Canada is disproportionately high relative to the national average. Graham puts it plainly: "facebook is enormously popular ... it has a particularly high penetration um out in nova scotia pei newfoundland and new brunswick".
Statista data from 2024 confirms the picture: Facebook penetration in PEI and Newfoundland runs around 81–82%, Nova Scotia around 75–79%, all well above the Canadian average of roughly 61%. For trades recruitment in the region, ignoring Facebook in favor of channels built around national-average demographics is leaving candidates on the table.
The platform mix matters for any employer doing regional hiring in Atlantic Canada. A national job board reaches everyone; a regional Facebook presence reaches the people who are actually there.
The companies worth knowing
If you are an estimator, project manager, or contractor spending hours on manual quantity take-offs, Construction AI is the product Graham built for you. The platform converts PDF drawing packages into structured data — quantities, MasterFormat codes, earthworks geometry, CAD exports — so your team's time goes to analysis and pricing rather than counting. They are active on LinkedIn, Facebook, and YouTube.
If you are a contractor with a surge-hiring problem or a hard-to-fill skilled trade, Blueforce Logistics places tradespeople on contract across Canada, covering all travel logistics so the worker arrives ready to work.
If you are a Maritime tradesperson looking for contract work outside the province, Blueforce recruits via Indeed and Facebook and covers the full cost of getting you there.
The two businesses Graham is running answer a version of the same question: where is skilled time being wasted, and what's the fix? For estimators, it's software. For tradespeople sitting in a province while work exists somewhere else, it's a logistics company willing to handle the friction. Neither business is a coincidence.
Guest: Jeff Graham, Co-founder, Construction AI and Blueforce Logistics. Episode 8 of the Atlantic Construction Podcast. Watch the full episode. Facebook penetration data sourced from Statista 2024 provincial social-media survey.