Rudus Brings AI-Powered Estimation to Construction's Most Overlooked Trade

Summary

Rudus, a Y Combinator P26 startup founded by Rishi Pankhaniya and Sahil, launched on Hacker News with an AI-powered takeoff and estimation platform specifically designed for concrete subcontractors. The software automates the labor-intensive process of measuring and quantifying materials from concrete plan sheets—work that currently requires senior estimators to manually trace structures, measure dimensions, and build Excel spreadsheets with 300+ line items, a process that can take weeks or months.

The founders identified a critical gap in the construction software market: existing AI takeoff tools were built for general contractors and treat concrete as a checkbox, missing the specialized workflows that concrete estimators use to price work. Rudus takes a different approach, operating as a "copilot" rather than a black box replacement. When estimators upload structural PDFs, the platform uses proprietary computer vision models to auto-classify sheets, detect concrete elements, follow cross-references across drawing sets, and expand each element into full assembly line items with concrete, formwork, and rebar calculations. Estimators maintain full control, reviewing and overriding outputs before export to their existing workflows—a critical feature since they stake million-to-billion-dollar bids on these numbers.

Rudus's competitive advantages stem from its laser focus on concrete and its methodology of building alongside estimator workflows rather than replacing them. The company trains multiple proprietary computer vision models directly on customer data, with each customer interaction feeding back into model training to improve per-client accuracy over time. This approach directly addresses the trust barrier that has caused previous AI estimation tools to fail in construction, where accuracy and accountability are non-negotiable.

• The copilot approach prioritizes estimator control and verifiability over fully autonomous outputs, addressing the trust barrier that killed previous AI tools
• Targets concrete subcontractors whose estimation bottleneck prevents them from bidding on available work

Editorial Opinion

Rudus exemplifies the power of deeply specialized AI in underserved markets where incumbent solutions have stagnated. Rather than chase the risky goal of fully autonomous takeoff generation—which would require near-perfect accuracy that current models can't deliver—the founders built a tool that augments human expertise. This "human-in-the-loop" approach is pragmatic and likely to drive adoption in a risk-averse industry where a single estimating error can be catastrophic. If they execute well, Rudus could unlock billions in trapped productivity in construction.