The vertical AI that reads a blueprint like a veteran estimator - and turns HVAC takeoffs into quotes in a fraction of the time.
REBAR — the wordmark set against the plan sets its AI was built to read. Named for the steel bars inside concrete, construction's literal backbone.
Behind every commercial building sits a document most people never see: a plan set dense with ductwork, diffusers, VAV boxes and equipment schedules. For decades, quoting the HVAC on that building meant an estimator combing through the sheets by hand - counting equipment, matching it to rooms, and assembling a bill of materials one line at a time. It was slow, careful, and easy to get wrong.
Rebar, a New York company founded in October 2024, was built to compress that work. Its software is hyper-specialized for one job: HVAC takeoffs. Proprietary computer vision models read construction blueprints and spec books, identify and categorize the equipment, count it, and generate a bill of materials and quote. The company says the process runs 60 to 70 percent faster than traditional methods, and that some manual tasks are compressed by more than 90 percent.
The strategy is narrow on purpose. Rather than serve every trade, Rebar goes deep on one - a bet its founders argue is the whole point. "Most companies either concentrate on other trades or try to cover all of them at once, which ultimately means they serve none particularly well," says CEO and co-founder Evan Brown.
That focus has drawn attention. In March 2026, Rebar closed a $14 million Series A led by Prudence, and reported doubling its annual recurring revenue in the first six weeks of the year.
Rebar sells to commercial HVAC equipment suppliers and manufacturer representative firms - and, more precisely, to the inside estimators and sales engineers inside them.
These are the people who turn an architect's plan set into a priced quote. When a project goes out to bid, they have limited time to review tens of thousands of data points across blueprints and specification books, count every piece of equipment, and get a number back before a competitor does. Miss the window, or miscount the equipment, and the bid is lost.
Rebar's answer is to automate the counting and marking, not the judgment. The estimator still owns the quote; the software removes the part of the job that is repetitive, error-prone and slow. Early users, the company says, are bidding two to three times more often and winning more of those bids - the metric that actually pays the bills.
"Rebar is delivering tools that are compressing manual task time by over 90%." Jordan Viniar — Prudence
Upload the blueprints and spec books. Rebar's models extract tens of thousands of data points from the project.
Computer vision identifies, categorizes and counts HVAC equipment - AI equipment marking with properties, plus automatic room search.
AI addendum summaries flag what changed between revisions so nothing slips between drawing sets.
Rebar assembles a bill of materials and generates the quote - hours of work compressed into minutes.
Automatically identifies HVAC equipment on plan sets and assigns properties, replacing hours of manual marking.
Locates and matches equipment to rooms across the drawing set so nothing is double-counted or missed.
Synthesizes project modifications so estimators see exactly what changed between revisions.
Tracks bid volume and performance, giving managers a read on throughput and outcomes.
Surfaces specification insights across the market to sharpen how firms position their bids.
Per-organization data isolation and NDA project protection - built for confidential bids.
"Most companies either concentrate on other trades or try to cover all of them at once, which ultimately means they serve none particularly well." Evan Brown — CEO & Co-Founder, Rebar
Construction estimating has no shortage of software. Horizontal takeoff and estimating tools - PlanSwift, STACK, Bluebeam Revu, Autodesk Takeoff - serve many trades at once, and newer AI entrants like Beam AI are pushing on the same problem. Rebar's wager is that the mechanical room rewards specialists. By committing to HVAC, it aims to be the tool an estimator can't imagine switching away from, then to repeat that depth in adjacent trades.
That places Rebar in the wave of "vertical AI" companies applying computer vision and agents to document-heavy, unglamorous workflows that big platforms overlook. The founding team's own history - Evan Brown worked summers for his uncle's commercial HVAC distributor before becoming an estimator and sales engineer - is the kind of domain grounding that vertical AI tends to reward.
Evan Brown and Andrew Schwartz start the company in New York to bring AI to HVAC estimating.
AI equipment marking, automatic room search and addendum summaries reach commercial HVAC suppliers.
Named an AHR 2026 Innovation Awards winner; annual recurring revenue doubles in the first six weeks of the year.
Prudence leads the round, with Zero Infinity Partners, Founder Collective, Villain Capital and Optimist Ventures. Next stops: plumbing and electrical.
Rebar makes AI estimating software for commercial HVAC equipment suppliers. Its computer vision reads construction blueprints and spec books, identifies and counts HVAC equipment, and generates a bill of materials and quote - cutting takeoff time by roughly 60-70%.
Rebar was founded in October 2024 by CEO Evan Brown, a former HVAC estimator, and co-founder Andrew Schwartz. It is based in New York.
Rebar raised a $14 million Series A in March 2026, led by Prudence with participation from Zero Infinity Partners, Founder Collective, Villain Capital and Optimist Ventures.
Commercial HVAC equipment suppliers and manufacturer rep firms - specifically the inside estimators and sales engineers who prepare quotes. Rebar reported about 40 customers at its Series A, seven of whom are also investors.
Most estimating tools serve many trades. Rebar specializes only in HVAC, going deep on that trade's equipment and workflows, with plans to expand into electrical and plumbing through additional AI agents.
Product demos and interviews: Rebar publishes a "Book a Demo" walkthrough on its site at withrebar.ai. No official YouTube channel was confirmed at publication - check the website's demo booking for the latest product video.