PLAN0 ✓

PLAN0 is a Y Combinator-backed (Spring 2026 batch) AI platform that turns architectural plans into construction cost estimates and analytics in minutes, using computer vision to parse floor plans, elevations, and other drawing types directly from PDF files. Rather than manually re-keying quantities and applying unit costs by hand -- a process that traditionally takes one to two weeks per project -- PLAN0's pitch is compressing that into roughly 30 minutes, automating the parsing, material extraction, and cost-breakdown steps that consume the bulk of a traditional preconstruction estimate.

The platform's own framing -- "building the Bloomberg of construction" -- signals its broader ambition beyond a single-project estimating tool: PLAN0 aggregates historical and real-time project cost data across geographies to power market-observable cost analytics, not just a one-off estimate for whatever plan is uploaded. According to the company's own Y Combinator launch materials, more than $20 billion of projects are already running through the platform.

Core capabilities include 24-month cost forecasting with confidence bands (rather than a single point estimate), real-time "what-if" scenario modeling for value-engineering decisions, and 3D model visualization generated directly from the parsed 2D plans. A demo project referenced in PLAN0's own materials identified $5.8 million in potential savings opportunities -- illustrative of the value-engineering use case the platform is built around, though prospective users should treat any single demo figure as illustrative rather than a guaranteed outcome for their own project.

The target customer base spans three distinct roles: developers use it for investment analysis and preconstruction routing; investors use it for cost validation, portfolio benchmarking, and stress-testing; general contractors use it for early business-development estimating and procurement planning. That three-sided positioning is deliberate -- PLAN0 is pitching itself as shared cost-intelligence infrastructure across a deal's stakeholders, not a tool built for one role in isolation.

PLAN0 was part of a notably large Y Combinator batch focused specifically on real estate and construction technology -- press coverage of that cohort described it as featuring 126 companies dominated by AI back-office tools targeting the industry's costliest operational bottlenecks, positioning PLAN0 within a broader wave of well-funded, technically sophisticated startups all attacking similar preconstruction and cost-intelligence problems around the same time, rather than as an isolated point solution.

PLAN0's founding team combines real estate, engineering, and data-science backgrounds specifically: CEO Arash Barati brings 12 years in real estate private equity, development, and lending across more than $7 billion of projects; CTO Shervin Barati spent eight years as an engineering manager at Apple; and co-founder Dimitris Pagonakis previously worked at Point72 and Citadel on data and analytics. That combination of domain expertise (real estate) and technical depth (computer vision, quantitative finance) is a meaningful credibility signal for a tool whose core promise is turning unstructured drawings into trustworthy financial data.

Editor’s takeBIM•STACK editor's take: PLAN0 tackles a genuinely painful, well-understood bottleneck -- early-stage cost estimation from architectural drawings -- with a founding team whose backgrounds map directly onto the problem (real estate finance plus computer vision). The $20B-in-projects claim and the founders' credentials are the strongest signals of traction at this early stage; the honest caveat is that PLAN0 doesn't publish pricing, so any firm evaluating it should book a demo expecting a sales conversation rather than instant self-serve access, and should sanity-check AI-generated estimates against an experienced estimator before using them for binding financial decisions.

PLAN0's core workflow -- uploading a PDF plan set and receiving a cost breakdown -- is designed to require minimal training, since the computer-vision parsing and cost-mapping happen automatically rather than through a manual takeoff process a user has to learn. The main skill investment for teams is learning to interpret and trust the confidence bands and scenario-modeling outputs, and to know when a generated estimate needs a human estimator's review versus when it's reliable enough to act on directly.

Because PLAN0 is still an early-stage, demo-gated product, the practical onboarding experience likely involves direct vendor support during initial adoption rather than a fully self-serve tutorial path -- firms evaluating it should expect a guided first project rather than an unassisted trial.

Teams already comfortable interpreting a traditional cost-estimating output (a breakdown by CSI division or building system) will find PLAN0's output format familiar, since the platform is generating the same kind of structured cost data an estimator would produce manually -- the novelty is in how quickly and automatically that structure gets populated from an uploaded plan set, not in a new way of reading or acting on the resulting numbers.