A concrete estimator somewhere in the Southwest—let's call him Mike, though his name could just as easily be Carlos or Jennifer—is clicking through hundreds of pages of foundation plans in Bluebeam, measuring slabs and footings by hand. The job is due Friday. Mike's backlog has three more plan sets behind it. By the time he finishes, the general contractor will have already picked someone else.
This is the grind that Rudus wants to break.
The San Francisco startup, which went through Y Combinator's Spring 2026 batch, has built an AI-powered takeoff platform designed to read construction drawings and calculate concrete quantities automatically. Upload a plan set, the company says, and what once took days now takes hours. Estimators review the numbers, tweak what needs adjusting, and move on to the next bid. The promise: submit three to five times more proposals, potentially boost revenue per estimator by half.
Those figures come straight from Rudus's marketing materials as of mid-May 2026. Independent validation? Not yet public. Customer logos plastered across the homepage? Absent. It's early—perhaps earlier than the founders might prefer if they had a choice—but the product is live, and contractors are kicking the tires.
The Pitch: Automate the Tedious, Bid More Jobs
Rudus isn't trying to reinvent estimating. It's trying to eliminate the part that makes seasoned estimators want to throw their laptops out a window: the endless clicking, measuring, cross-checking against specs buried twelve sheets deep in a PDF.
The platform detects concrete elements—slabs, columns, walls, footings, rebar—and calculates quantities automatically. Estimators can adjust the numbers in-browser, then either price the job inside Rudus or export everything to HCSS HeavyBid, Sage Estimating, B2W Estimate, or Excel. Multiple users can work on the same plan set simultaneously, no emailing files back and forth or waiting for someone to release a seat license.
One feature stands out, at least on paper: "Ask the plan set anything." It's a natural-language search tool that pulls answers from plans, details, and specs, with citations linked back to the source. The idea is to keep estimators moving instead of losing twenty minutes hunting for a callout on sheet 47.
Whether contractors actually use it that way, or whether it becomes one of those features that looks great in a demo and collects dust in practice, remains an open question.
Fourteen Days to Deploy (They Say)
Rudus claims onboarding takes two weeks. Day one, you upload a sample sheet. Day two, the system ingests past jobs to calibrate itself. Day seven, team training—under an hour per session. Day fourteen, you're exporting live bids.
It's a tidy timeline. Tidy enough to make you wonder if reality is quite that clean, especially for contractors whose IT infrastructure consists of a guy named Dave who also handles jobsite logistics. Pricing isn't public; interested firms go through a demo request process, which is standard for B2B software but also conveniently keeps competitive intelligence under wraps.
The founders—Sahil Goel and Rishi Pankhaniya—are both in San Francisco, both came up through big tech. Goel logged time at Amazon, NASA's Jet Propulsion Laboratory, and Northrop Grumman before this. Pankhaniya worked at Airbnb (where he was reportedly the first new grad engineer hired in over two years) and Apple. They met at Y Combinator, launched Rudus in 2026, and now occupy that peculiar startup stage where the team size is listed as "two to ten employees" on LinkedIn.
Y Combinator is backing the company. A LinkedIn post announcing Rudus racked up 578 likes and 200 comments within 15 hours, according to a snapshot from mid-May 2026. Social proof, or at least the appearance of it.
AI Comes for the Takeoff
Rudus is hardly alone. The construction tech space has spent the past few years watching AI migrate from "interesting experiment" to "expected feature."
A December 2025 survey by Dodge Construction Network and CMiC found that 85 percent of contractors anticipate AI will cut time spent on repetitive tasks. Three-quarters think it'll help them learn from historical data. Autodesk, in a March 2026 blog post, framed the shift as estimators evolving into strategic advisors rather than button-clickers—a tactful way of saying the tedious part might soon be handled by software.
Digital takeoffs already offer speed gains. Bluebeam's 2026 guide pegs digital workflows at three to four times faster than manual methods, once teams adjust. AI-native platforms are pushing that further. Togal markets "up to 98 percent accuracy" in its takeoff engine. Autodesk offers Forma Takeoff for 2D and 3D quantification. STACK Construction Technologies rolled out AI-centric features around the World of Concrete 2024 conference. Sage Estimating's 2026 brochure mentions "AI-powered takeoff tools" in its ecosystem, though the details are vague. XBuild targets smaller residential concrete jobs—driveways, patios, the stuff that keeps a two-person crew busy all summer.
Rudus is aiming at a specific niche: concrete and rebar subcontractors who need to churn through bids fast enough to stay in the game. Whether that focus is strategic brilliance or a constraint the company will outgrow depends on how quickly they can prove the product works at scale.
What's Not There Yet
No customer case studies. No logos. No independent audits of those 70-percent-faster claims. No disclosed funding beyond Y Combinator. The product demo posted in late April shows a clean interface, but real-world adoption data remains locked behind demo requests and sales calls.
For contractors who've spent the past decade hearing about how software will revolutionize their business—only to wind up with clunky interfaces and support tickets that go unanswered for days—this absence of proof is worth noting. Rudus might be the real deal. It might also be another well-funded pivot waiting to happen.
The tool is live. Contractors are testing it. The founders have credibility on paper, backing from one of the most respected startup accelerators in the world, and a product that at least theoretically solves a painful, expensive problem.
Now comes the harder part: turning demo requests into deployed seats, marketing claims into customer testimonials, and a clever idea into a business that concrete contractors actually trust with their bids. Mike—or Carlos, or Jennifer—is still clicking through Bluebeam. Whether Rudus can get them to stop is the question the next twelve months will answer.
